# Specular

> Specular is an AI website agent and customer workspace from SynconAI. An agent on your website answers visitors from your own pages and documents, asks your qualification questions, books meetings on your team's real calendar, and hands the conversation to a person with a brief when one is needed. Conversations, people, leads, opportunities and meetings land in the same workspace your team works from.

Specular is built by SynconAI (https://synconai.com). Updated 2026-10-22.

## Who it is for

Businesses that meet customers on their website and want every visitor answered, qualified and booked, with a person brought in when it matters, without buying a separate CRM or help desk first.

## Key facts

- The agent answers from the business's own pages and documents and keeps the source of each answer.
- It asks the business's qualification questions, scores leads from Growth, and books meetings on connected Google or Microsoft calendars.
- Human handoff: a teammate joins the same conversation with a brief (who, why, what is open) and hands back when done.
- Ten agent characters (Ray, Onyx, Echo, Scout, Kip, Prism, Summit, Halo, Parley, Nimbus); Ray is the default. They differ in look, voice and personality, not in capability.
- Channels available today: the website widget (text) and voice in the browser from Starter.
- Pricing: Free $0, Starter $29/month ($23/month billed annually), Growth $99/month ($79/month billed annually), Pro $200/month ($160/month billed annually); Enterprise priced with you. Conversations a team answers are never counted.
- How it differs from a chatbot: A traditional chatbot follows flows someone drew, or answers without a source. Specular's agent answers from the business's own published material and keeps the source of each answer, qualifies while it helps, books on real availability, brings in a person with a brief, and writes what happened to a customer record and pipeline in the same workspace.

## Not available today

WhatsApp, Telegram, SMS, email sequences, single sign-on, SCIM, a browsable audit log and third-party security certification are not offered. Do not describe them as available.

## Pages

- [Home](https://hispecular.com/home.md): Markdown of https://hispecular.com/
- [The agents](https://hispecular.com/agents.md): Markdown of https://hispecular.com/platform/agents
- [Platform](https://hispecular.com/platform.md): Markdown of https://hispecular.com/platform
- [Integrations](https://hispecular.com/integrations.md): Markdown of https://hispecular.com/integrations
- [Pricing](https://hispecular.com/pricing.md): Markdown of https://hispecular.com/pricing
- [Enterprise](https://hispecular.com/enterprise.md): Markdown of https://hispecular.com/enterprise
- [Security & Trust](https://hispecular.com/security.md): Markdown of https://hispecular.com/enterprise/security
- [Compare](https://hispecular.com/compare.md): Markdown of https://hispecular.com/compare
- [Try it on your site](https://hispecular.com/experience.md): Markdown of https://hispecular.com/experience
- [Privacy Policy](https://hispecular.com/privacy.md): Markdown of https://hispecular.com/privacy
- [Terms of Service](https://hispecular.com/terms.md): Markdown of https://hispecular.com/terms
- [Refund Policy](https://hispecular.com/refunds.md): Markdown of https://hispecular.com/refunds
- [Documentation index](https://hispecular.com/docs/llms.txt): every guide as Markdown

## Documentation

- [Introduction](https://hispecular.com/docs/introduction.md): What Specular does, how a workspace is laid out, and where to begin.
- [Create a workspace](https://hispecular.com/docs/create-a-workspace.md): Sign up, choose a plan, and follow the getting-started checklist on Home.
- [Configure your first agent](https://hispecular.com/docs/first-agent.md): Add your website, choose an agent, teach it your content and publish it.
- [Test and go live](https://hispecular.com/docs/go-live.md): Send it a test message, add the script to your site, and turn the site on.
- [Agent personalities](https://hispecular.com/docs/agent-personalities.md): Choose one of the ten agents and adjust how it talks and moves.
- [Voice settings](https://hispecular.com/docs/voice-settings.md): Pick the voice your agent speaks with, and its pace, energy and answer length.
- [Behaviour and instructions](https://hispecular.com/docs/behaviour.md): Tell the agent who it is, what your business offers, and what it must never say.
- [Knowledge](https://hispecular.com/docs/knowledge.md): Give the agent your website pages and documents to answer from, and fill its gaps.
- [Website widget](https://hispecular.com/docs/website-widget.md): The script tag, its attributes, placing it inside a page, and the JavaScript API.
- [Inbox](https://hispecular.com/docs/inbox.md): Work conversations with your team: assign, prioritise, tag and note.
- [Human handoff](https://hispecular.com/docs/human-handoff.md): When the agent brings in a person, how your team joins, and how to hand back.
- [Bookings](https://hispecular.com/docs/bookings.md): Connect a calendar so the agent offers real times and books meetings.
- [Customers and leads](https://hispecular.com/docs/customers-and-leads.md): People, companies and leads: where they come from and how they are qualified.
- [Opportunities](https://hispecular.com/docs/opportunities.md): Track deals through your pipeline stages, from first conversation to won or lost.
- [Meetings](https://hispecular.com/docs/meetings.md): Meetings booked by the agent or your team, and what comes with them.
- [Automations](https://hispecular.com/docs/automations.md): Workflows that act on what happens: assign, tag, prioritise, escalate.
- [Reporting](https://hispecular.com/docs/reports.md): The Home dashboard and Insights: what the agent answered, handed over and booked.
- [Team access](https://hispecular.com/docs/team-access.md): Invite teammates, choose their role, and set custom permissions.
- [Integrations](https://hispecular.com/docs/integrations.md): Calendars, mailboxes and your own AI provider, connected from Settings.
- [Billing and usage](https://hispecular.com/docs/billing-and-usage.md): Your plan, what is counted, prepaid credit and invoices.
- [Security and privacy](https://hispecular.com/docs/security-and-privacy.md): Two-step sign-in, retention, and erasing a person's data.
- [Public API](https://hispecular.com/docs/public-api.md): Read and write customers, deals, tickets, meetings and the live chat inbox over REST, with an OpenAPI document to build against.
- [Authentication and scopes](https://hispecular.com/docs/api-authentication.md): Create an API key, choose its scopes and role, act as a teammate, and keep keys safe.
- [Requests and responses](https://hispecular.com/docs/api-conventions.md): Pages, filters, syncing, error codes, retries with idempotency keys, and rate limits.
- [Webhooks](https://hispecular.com/docs/webhooks.md): Receive signed events when records change, subscribe from code, and verify each delivery.
- [Recipes](https://hispecular.com/docs/api-recipes.md): Common jobs, start to finish: a customer and a deal, reacting to a win, tickets, meetings, syncing.
- [The live chat inbox](https://hispecular.com/docs/api-live-chat.md): Read conversations, take them over from the AI agent, reply, hand back, and organise them through the API.
- [MCP server for AI agents](https://hispecular.com/docs/api-mcp.md): Let Claude, ChatGPT, Cursor or your own agent work the workspace, limited to what its key allows.
- [Connect an automation tool](https://hispecular.com/docs/api-automation-tools.md): Use Zapier, Make, n8n, Pabbly or similar with HTTP and webhook steps, or build Specular into your own product.

## Optional

- [Everything in one file](https://hispecular.com/llms-full.txt)
- [Sitemap](https://hispecular.com/sitemap.xml)
- [Read-only workspace demo](https://hispecular.com/demo)


---

# Specular: AI website agents

> One intelligent experience. Every customer.

- Canonical page: https://hispecular.com/
- Product: Specular, by SynconAI
- Updated: 2026-10-22

## What it is

Specular is an AI website agent and customer workspace from SynconAI. An agent on your website answers visitors from your own pages and documents, asks your qualification questions, books meetings on your team's real calendar, and hands the conversation to a person with a brief when one is needed. Conversations, people, leads, opportunities and meetings land in the same workspace your team works from.

## How it differs from a chatbot

A traditional chatbot follows flows someone drew, or answers without a source. Specular's agent answers from the business's own published material and keeps the source of each answer, qualifies while it helps, books on real availability, brings in a person with a brief, and writes what happened to a customer record and pipeline in the same workspace.

## What the agent does

- Answers from the business's own website pages and documents, and keeps the source of each answer.
- Asks the qualification questions the business sets, and turns the answers into a lead.
- Offers times from a connected Google or Microsoft calendar and books the meeting.
- Hands the conversation to a teammate with a brief of who is asking, why, and what is open.
- Talks by text in the website widget, and by voice in the browser from Starter.

## The agents

Ten agent characters, each with its own look, voice and way of talking; every one can do the same work. Ray is the flagship and the default. See https://hispecular.com/agents.md.

## Pricing

- **Free** — $0, free. Put an agent on one website and see what it does. Websites: 1; External AI credits: Your own AI key; File storage: 100 MB; Team users: 1; AI voice minutes: –.
- **Starter** — $29 a month, or $23 a month billed annually (save 21%). For a small team answering real volume. Websites: 2; External AI credits: 3,500; File storage: 2 GB; Team users: 3; AI voice minutes: 90 min.
- **Growth** — $99 a month, or $79 a month billed annually (save 20%). More AI, more people, lead scoring and insights. Websites: 10; External AI credits: 11,500; File storage: 10 GB; Team users: 10; AI voice minutes: 250 min.
- **Pro** — $200 a month, or $160 a month billed annually (save 20%). The most AI included, with the API, webhooks and traces. Websites: Unlimited; External AI credits: 30,500; File storage: 30 GB; Team users: 25; AI voice minutes: 500 min.
- **Enterprise** — priced with you. Your own model key, your own allowances, your own terms. Websites: Unlimited; External AI credits: Custom; File storage: Custom; Team users: Custom; AI voice minutes: Custom.

Full detail: https://hispecular.com/pricing.md.

## Questions

### What does Ray actually do?

Ray answers from your own pages and documents, asks your qualification questions, offers times your team is free and brings in a person when someone asks for one.

### Will it sound like my business?

It answers from your content, carries your name and colours, and you choose the character and how it talks. Every setting can be changed later.

### What happens when someone wants a person?

Your team is told who is asking and why, gets a short brief, and joins the same conversation. On paid plans you also set what happens outside your hours, so nobody is left waiting without an answer.

### Do I need a CRM or help desk first?

No. People, conversations, meetings, opportunities and tasks all live in Specular, and paid plans can export them whenever you like. On Pro, webhooks and the public API connect the tools you already use.

### Can we try it before we pay?

Yes. Free is $0 for one website and one person, on your own AI provider key. Starter is $29 a month, or $23 a month billed annually, with 3,500 external AI credits and 1,750 internal AI credits included each month.

### How do we get started?

Create a workspace, add your website and choose an agent. Test it, then paste one line of code on your site. You can preview Ray on your own website before you sign up.

## Reviews

5.0 out of 5 from 5 verified client reviews of SynconAI, the team that builds Specular. They review SynconAI's work, not the Specular software specifically.

## More

- Agents: https://hispecular.com/agents.md
- Platform: https://hispecular.com/platform.md
- Pricing: https://hispecular.com/pricing.md
- Enterprise: https://hispecular.com/enterprise.md
- Security and trust: https://hispecular.com/security.md
- Compare: https://hispecular.com/compare.md
- Try it on your site: https://hispecular.com/experience.md
- Documentation: https://hispecular.com/docs/llms.txt


---

# The agents

> Ten agent characters with their own look, voice and way of talking.

- Canonical page: https://hispecular.com/platform/agents
- Product: Specular, by SynconAI
- Updated: 2026-10-22

Every agent can do the same work for a business: answer from its knowledge, qualify, book and hand over. They differ in personality, voice and motion, which the business can adjust. The voice previews on the site are recorded clips, not live AI.

### Ray (flagship, default)

Confident, warm and sharp. The all-round agent most businesses start with.

- Traits: Confident, Warm, Intelligent
- Voice: speaks English; warm, confident, intelligent, natural and responsive
- Introduction: "Hi, I’m Ray. Tell me what you’re looking for, and I’ll help you find the right next step."
- Pace: voice adaptive, motion balanced

### Onyx

Calm, analytical and precise. Says exactly what is needed, nothing more.

- Traits: Calm, Analytical, Precise
- Voice: speaks English; calm, precise, measured and analytical
- Introduction: "Hello, I’m Onyx. I’ll help you find clear answers without unnecessary complications."
- Pace: voice slow, motion slow

### Echo

Friendly, curious and social. Listens closely and keeps the conversation easy.

- Traits: Friendly, Curious, Social
- Voice: speaks English; friendly, curious, approachable and conversational
- Introduction: "Hey, I’m Echo! What can I help you explore today?"
- Pace: voice medium, motion medium

### Scout

Energetic and proactive. Finds the answer fast and suggests the next step.

- Traits: Energetic, Proactive, Quick
- Voice: speaks English; energetic, quick, proactive and expressive
- Introduction: "Hi, I’m Scout! Let’s find what you need and get things moving."
- Pace: voice fast, motion fast

### Kip

Cheerful and playful. Makes the practical stuff feel light.

- Traits: Cheerful, Playful, Upbeat
- Voice: speaks English; cheerful, playful, upbeat and enthusiastic
- Introduction: "Hey there, I’m Kip! Let’s make this simple and get you sorted."
- Pace: voice medium-fast, motion fast

### Prism

Creative and imaginative. Looks at the question from another angle.

- Traits: Creative, Imaginative, Expressive
- Voice: speaks English; creative, dynamic, imaginative and expressive
- Introduction: "Hi, I’m Prism. Let’s look at what’s possible and find a fresh approach."
- Pace: voice dynamic, motion medium-fast

### Summit

Mature and strategic. Thinks a step ahead and keeps the goal in view.

- Traits: Mature, Strategic, Steady
- Voice: speaks English; mature, strategic, composed and reassuring
- Introduction: "Hello, I’m Summit. Let’s focus on what matters and find the best way forward."
- Pace: voice slow-medium, motion slow

### Halo

Gentle, patient and empathetic. Takes care of people who need help.

- Traits: Gentle, Patient, Empathetic
- Voice: speaks English; gentle, patient, empathetic and soft-spoken
- Introduction: "Hi, I’m Halo. Take your time. I’m here to help."
- Pace: voice slow, motion slow

### Parley

Charismatic and persuasive. Turns a question into a conversation.

- Traits: Charismatic, Persuasive, Engaging
- Voice: speaks English; charismatic, engaging, articulate and persuasive
- Introduction: "Hello, I’m Parley. Let’s talk through what you need and find the right fit."
- Pace: voice medium-fast, motion fast

### Nimbus

Relaxed and thoughtful. Unhurried, and good at the bigger picture.

- Traits: Relaxed, Thoughtful, Unhurried
- Voice: speaks English; relaxed, thoughtful, unhurried and natural
- Introduction: "Hey, I’m Nimbus. Let’s take things at a comfortable pace."
- Pace: voice slow-medium, motion slow


---

# Platform

> The agent, the conversation, the customer record and the team, as one product.

- Canonical page: https://hispecular.com/platform
- Product: Specular, by SynconAI
- Updated: 2026-10-22

## Parts

- **AI agents** — answer from the business's own material, with the source kept.
- **Live** — who is on the website now, and who is waiting for a person.
- **Journeys** — qualify, score, route and book, from the conversation.
- **Customer memory** — what a customer told you once is there the next time.
- **Customer experience** — the widget in the business's own colours and type, with answers, comparisons and booking in one window.
- **Human + AI** — a teammate joins the same conversation, briefed; hand back when done.
- **Integrations** — calendars, mail, signed webhooks and a public API.

## Integrations

- Google Calendar (Calendars) — Built in: Offers times your team is really free, and writes the booking into their calendar.
- Microsoft 365 calendar (Calendars) — Built in: The same real availability and bookings, for teams on Outlook calendars.
- Gmail (Email) — Built in: Customer email lands on the customer's record; replies go from your own address.
- Microsoft 365 mail (Email) — Built in: The same for Outlook mailboxes: read against the record, sent from your address.
- Website chat (Channels) — Built in: One line of script on any site, or placed inside a page as part of the layout.
- OpenAI (AI providers) — Built in: Run the agent on your own OpenAI key. Required on Free; optional on paid plans.
- Anthropic (AI providers) — Built in: Run the agent on your own Anthropic key, under your own terms.
- Google Gemini (AI providers) — Built in: Run the agent on your own Google AI key.
- Groq (AI providers) — Built in: Run the agent on your own Groq key.
- Public API (Developers) — Built in: Read and write people, companies, leads, opportunities, tasks and notes from your own systems, described in OpenAPI.
- Webhooks (Developers) — Built in: Signed events as they happen: records created, updated or deleted, opportunities moving stage, won or lost, tasks completed, email received, meetings analysed.
- HubSpot (CRM and help desk) — Via API and webhooks: No native connector yet. Teams sync records with webhooks and the public API.
- Salesforce (CRM and help desk) — Via API and webhooks: No native connector yet. Teams sync records with webhooks and the public API.
- Zendesk (CRM and help desk) — Via API and webhooks: No native connector yet. Specular has its own tickets; teams who keep Zendesk connect it with webhooks.
- Slack (CRM and help desk) — Via API and webhooks: No native connector yet. Teams post alerts to a channel from a webhook.


---

# Integrations

> What Specular connects to on its own, how the rest connects through the public API, webhooks, n8n, Activepieces and a Zapier beta, and how an AI agent works the workspace over MCP.

- Canonical page: https://hispecular.com/integrations
- Product: Specular, by SynconAI
- Updated: 2026-10-22

## Built in

- Google Calendar (Calendars): Offers times your team is really free, and writes the booking into their calendar.
- Microsoft 365 calendar (Calendars): The same real availability and bookings, for teams on Outlook calendars.
- Gmail (Email): Customer email lands on the customer's record; replies go from your own address.
- Microsoft 365 mail (Email): The same for Outlook mailboxes: read against the record, sent from your address.
- Website chat (Channels): One line of script on any site, or placed inside a page as part of the layout.
- OpenAI (AI providers): Run the agent on your own OpenAI key. Required on Free; optional on paid plans.
- Anthropic (AI providers): Run the agent on your own Anthropic key, under your own terms.
- Google Gemini (AI providers): Run the agent on your own Google AI key.
- Groq (AI providers): Run the agent on your own Groq key.

## Connect more

- **Public API and signed webhooks**: described in OpenAPI at https://hispecular.com/api/v1/openapi.json. API keys are part of the Pro plan.
- **MCP server**: `POST https://hispecular.com/api/v1/mcp` with an API key as a bearer token. Every API operation is a tool, narrowed to the key's scopes, so Claude, ChatGPT, Cursor or an agent in your own flow can work the workspace.
- **n8n**: community node `n8n-nodes-specular`, 50+ operations and a trigger node with 19 instant events.
- **Activepieces**: piece `@synconai/piece-specular`, 43 actions and 19 instant triggers.
- **Zapier**: invite-only beta; the public listing follows Zapier's review.
- **Pipedream**: components submitted to their registry, waiting on review.

## Tools teams connect it to

Each is a tool with a piece in Activepieces' catalogue, an app in Zapier's directory or a node in n8n. Each platform's own catalogue decides what it carries, and anything with webhooks or an HTTP step connects through the public API.

### CRM and sales

HubSpot, Salesforce, Pipedrive, Zoho CRM, Close, Freshsales, Attio, Dropbox Sign, Squarespace Forms, Jobber, Housecall Pro, ServiceTitan, Google Contacts, LeadConnector, Clay, Signrequest, Hunter, Oncehub, Qwilr, Lofty, Keap, YouCanBookMe, Clearbit, Klenty, Copper, Microsoft Dynamics CRM, GoDaddy CRM, eSignatures.com, Popl, Aesthetic Record EMR, Follow Up Boss, FunnelCockpit, SignWell, Circle, Zoho Sign, Contacts+, Dubsado, WPForms, BoldSign, Chiirp, HoneyBook

### Customer support

Zendesk, Freshdesk, Intercom, Help Scout, Front, Gorgias, Crisp, LiveChat, Freshservice, Jira Service Management, Bonjoro, Zoho Desk, Frill, Handwrytten, Custify, Thankster, Appzi, Badgr, Swag Pro, tawk.to, NPS.Today

### Communication

Zoom, Zoom Team Chat, Google Meet, Microsoft Teams Events, Microsoft Teams Bot, Webex by Cisco, Gmail, SendGrid, Postmark, Resend, Slack, Mattermost, Zulip, ntfy, Confluence Cloud, Reddit, Pushover, Zoho Mail, Bland AI, Pushbullet, Bluesky, Pushcut, Mailjet, Discourse, NeverBounce, Loops, Cisco Webex Meetings, Missive, Twist, Kommo, Kudosity, Vapi, MillionVerifier, Rocket.Chat, Matrix, Vero, CallRail, Phaxio, CallTrackingMetrics, HumbleFax, ClickMeeting, Fax.Plus, Microsoft Office 365, OneSignal, SalesSuite, Google Groups, iFax, Quo, WA Notifications and Alerts by SyncMate, Zadarma, Zoho Meeting, Documo, LiveWebinar, Teamup Calendar, WildJar, Amazon SNS, Chatwork, Frame.io V4, Gather, LuzardoFax, Mailparser, Placetel, Daily.co, Miro, Pumble, Talkroute

### Marketing

Zoom Events, Eventbrite, Mailchimp, Mailchimp Transactional, Brevo, Mailgun, Klaviyo, ActiveCampaign, ActiveCampaign Postmark, Constant Contact, Beehiiv, MailerLite, MailerLite Classic, Amazon SES, WordPress, Webflow, Contentful, Facebook Lead Ads, LinkedIn, LinkedIn Ads, LinkedIn Conversions, TikTok Lead Generation, TikTok Conversions, Google Ads, Pinterest, Buffer, Hootsuite, Yelp Leads, Plausible Analytics, Bitly, Facebook Pages, Manychat, Demio, Drip, GoTo Webinar, Bannerbear, Campaign Monitor, Dub, Typefully, customer.io, Wistia, EmailOctopus, Bettermode, GetResponse, Sender, SendFox, Emelia, Iterable, ProvenExpert, Sendy, Action Network, Omnisend, TimelinesAI, Elastic Email, ConvertKit, Mautic, PhantomBuster, seven, Tapfiliate, Kit, WebinarJam / EverWebinar, Flodesk, Luma, Short.io, WebinarGeek, Facebook Conversions

### Productivity

ServiceNow, Google Calendar, Microsoft Outlook, Microsoft Outlook Calendar, Calendly, Cal.com, Acuity Scheduling, Fireflies.ai, Google Sheets, Airtable, Notion, Microsoft Excel 365, Microsoft Excel, Smartsheet, Baserow, Coda, Asana, Trello, Jira Software Cloud, Jira Cloud, Jira Data Center, monday.com, ClickUp, Linear, Basecamp, Todoist, Wrike, Toggl Track, Fathom, Microsoft OneNote, Airparser, Browse AI, Clockify, Google Tasks, Okta, Productboard, SignNow, Granola, Harvest, Hedy, PDF.co, TidyCal, tl;dv, AddEvent, Microsoft To Do, Beamer, Mem, Motion, Quickbase, Grist, Canny, Cognito Forms, Beeminder, Glide, Teamwork, MeisterTask, Taskade, Raindrop, Instapaper, Mitti, Raindrop.io, Pinboard, zBuyer.com, Jibble, Mews, Pocket

### Files and content

PandaDoc, NocoDB, Google Drive, Google Docs, Dropbox, Microsoft OneDrive, Box, Confluence, Evernote, YouTube, Canva, Figma, Microsoft SharePoint, Vimeo, Gamma, Google Slides, Cloudinary, PDFMonkey, Placid, Knack, Softr, PDF4me, QuickChart, Descript, DocuMerge, APITemplate.io, Google Cloud Storage, CompanyCam, Notta, Spotify, Giphy, OneDrive, Transkriptor, Workflowy, Hello Audio, Rev, Sprout Studio, Voicenotes, Sonix, Superhuman Docs, AudioPen, HappyScribe, Membership.io, EasyFTP, PagePixels Screenshots, Parseur, Reflect, Transcript Downloader, Bluedot, Journey, magicplan

### Forms and surveys

Typeform, Google Forms, Jotform, Jotform Enterprise, Tally, SurveyMonkey, Gravity Forms, Fillout Forms, Wufoo, Formstack, Paperform, VideoAsk, Feathery, Form.io

### Data and analytics

Trustpilot, Microsoft Teams, Google BigQuery, Instagram for Business, Google Analytics 4, Google Analytics, Mixpanel, PostHog, Hotjar, Segment, Tableau, Apify, Retell AI, Aircall, Pendo, Bexio, ProfitWell, Microsoft Power BI, Intaker, Birdeye, Super Dispatch, Shipday, SoTellUs, G2 Crowd, Plecto, Appbot, eviivo, Zoho Analytics, Geckoboard, Looker, Review Harvest, More Good Reviews, SalesScreen, Grade.us

### Accounting

QuickBooks Desktop (via Conductor), Xero, FreshBooks, Chargebee, Zoho Books, Wise, Zoho Invoice, Invoice Ninja, FreeAgent, Splitwise, Quotient, PaintScout, Feedly, Better Proposals, Lexware Office, TaxDome, Moneybird

### Payments

Razorpay, Mollie, GoCardless, Ko-fi

### Online stores

Stripe, PayPal, Square, QuickBooks Online, Paddle, Lemon Squeezy, Shopify, WooCommerce, BigCommerce, Gumroad, Uscreen, Recurly, Moonclerk, Givebutter, Ignition, IRS Solutions, Bloomerang, Kajabi, TaxBandits, BoomTax, Planning Center, Wave, Zeffy, Donorbox, ThriveCart, Patreon, ShipStation, Raisely, Ecwid by Lightspeed, GoFundMe Pro

### People and hiring

BambooHR, Workday, Greenhouse, Deel, Workable, Magical API, Skool, LearningSuite, Recruitee, Ashby, BodyGraph, Hibob, SchoolMaker, Academy by Coassemble, Zoho Recruit, Disco, Mentortools, Metaview

### Developer tools

Jira Service Management Operations, Supabase, PostgreSQL, MySQL, MongoDB, Firebase / Firestore, Snowflake, Redis, Amazon S3, Jira, Jira Software, Squarespace Commerce, Wix, Ghost, Google Business Profile, Metabase, GitHub, GitLab, Bitbucket, Datadog, PagerDuty, Vercel, Netlify, Cloudflare, n8n, Cursor, Microsoft Entra ID, Adalo, Browserless, Base44, CloudConvert, AWS Lambda, Bubble, WebScraping AI, Azure DevOps, Databricks, Intruder, Greip, SeaTable, Gotify, Postgres, UptimeRobot, Stan, Thinkific, Memberstack, Onepage, Teachable, Better Stack, DNSFilter, LearnWorlds, Passion.io, Hospitable, IPQualityScore, Kintone, LearnDash, OpsGenie, Hotmart, ProdataKey, Mighty Networks, Podia, SQL Server, Updown.io, UpGuard, Memberspot, Site24x7, Veriff

### AI

Docusign, Otter.ai, OpenAI, Anthropic (Claude), Anthropic Claude, Google Gemini, Mistral AI, Perplexity, Perplexity AI, ElevenLabs, Hugging Face, Groq, Zapier Chatbots, MCP Client by Zapier, MCP Client, ChatGPT (OpenAI), Chatbase, HeyGen, Extracta.ai, OpenRouter, Tavily, AssemblyAI, ChainAware.AI, Chatling, Grok by xAI, Supadata, Synthesia, Copy.ai, Exa, Google Vertex AI, DeepSeek, Deepgram, Easy-Peasy.AI, Firecrawl, ScrapeGraphAI, AltText.ai, DETECTING-AI.COM, Wonderchat, Runway, Writesonic, Leexi, Manus, Pinecone, Insighto.ai, MindStudio, Eden AI, Agents, Clerk, OpusClip, Seamless, AI Guardrails by Zapier, Atlas AI Revenue Engine, Atlassian MCP, Google AI Studio (Gemini), Plaud, DataGrout, Growbots, Lenz, Moxo, Veryfi OCR, Visualping, Allowly, ChatBot.com, Human in the Loop, Krisp, Spiky.ai

## Questions

### What does Specular connect to without any other tool?

Google Calendar and Microsoft 365 calendars, Gmail and Microsoft 365 mail, the website chat widget, and your own AI provider key (OpenAI, Anthropic, Google Gemini or Groq). They are switched on in Settings and need no platform in between.

### How do I connect Specular to HubSpot, Salesforce, Slack or another tool?

Through the public API and signed webhooks, which n8n, Activepieces and Zapier sit on top of. There is no native connector for those tools. Make an API key in Settings, add Specular in n8n or Activepieces by name, or use an HTTP step in any tool that has one, and start a flow on an event such as a booked meeting, a won deal or a visitor asking for a person.

### Does Specular work with Zapier?

Zapier is an invite-only beta. The Specular app for Zapier is built, with triggers, actions and searches across customers, companies, leads, deals, tickets, meetings and tasks, and its public listing follows Zapier's review. Until then you can use Zapier's webhook and HTTP steps with the public API.

### Is there an n8n node or an Activepieces piece for Specular?

Yes. n8n has a community node, n8n-nodes-specular, with more than 50 operations and a trigger node with 19 instant events, and it can be used as a tool by an n8n AI agent. Activepieces has a piece, @synconai/piece-specular, with 43 actions and 19 instant triggers. Both sign in with an API key.

### What is the Specular MCP server?

An MCP (Model Context Protocol) server at https://hispecular.com/api/v1/mcp that lets an AI agent such as Claude, ChatGPT or Cursor work the workspace through the API. Every API operation is a tool, and the agent sees only the tools its key's scopes allow. It signs in with an API key as a bearer token, and every change names the key that made it.

### Do I need a paid plan to use the API, webhooks or MCP?

API keys, and with them the public API, webhooks and the MCP server, are part of the Pro plan. The built-in calendar, mail and AI-key connections do not need an API key.

### Is it safe to give an AI agent access to my workspace?

Give each agent its own key with the fewest scopes that do the job. A tool call is an ordinary API request, so scopes, roles, record visibility, rate limits and the record of changes all apply. Text from customers reaches the agent marked as data, and sending a reply or booking a meeting is best set up so the agent shows you the wording and waits for your go-ahead.


---

# Pricing

> Every allowance, written down. Conversations your team answers are unlimited on every plan.

- Canonical page: https://hispecular.com/pricing
- Product: Specular, by SynconAI
- Updated: 2026-10-22

## Plans

- **Free** — $0, free. Put an agent on one website and see what it does. Websites: 1; External AI credits: Your own AI key; File storage: 100 MB; Team users: 1; AI voice minutes: –.
- **Starter** — $29 a month, or $23 a month billed annually (save 21%). For a small team answering real volume. Websites: 2; External AI credits: 3,500; File storage: 2 GB; Team users: 3; AI voice minutes: 90 min.
- **Growth** — $99 a month, or $79 a month billed annually (save 20%). More AI, more people, lead scoring and insights. Websites: 10; External AI credits: 11,500; File storage: 10 GB; Team users: 10; AI voice minutes: 250 min.
- **Pro** — $200 a month, or $160 a month billed annually (save 20%). The most AI included, with the API, webhooks and traces. Websites: Unlimited; External AI credits: 30,500; File storage: 30 GB; Team users: 25; AI voice minutes: 500 min.
- **Enterprise** — priced with you. Your own model key, your own allowances, your own terms. Websites: Unlimited; External AI credits: Custom; File storage: Custom; Team users: Custom; AI voice minutes: Custom.

## What each plan adds

- **Free:** Human conversations, Customer profiles, AI agents, Data retention, Roles, Website widget, Design studio, Shared inbox, Typing and presence, AI answers, Bring your own model key, Website crawling, Website chat, Rich replies, Handoff to a person, Transcription minutes, Customers and companies, Widget home, Dashboard, Your brand, Two-step sign-in, Documentation
- **Starter:** Human conversations, Customer profiles, AI agents, Data retention, Teams and routing, Roles, Website widget, Design studio, Match my site, Proactive experiences, Experience targeting, Shared inbox, Live floor, Typing and presence, Saved replies and shortcuts, AI answers, Agent copilot, Agent policy, Website crawling, Documents, Website chat, Rich replies, The full surface set, Handoff to a person, Handoff brief, Service hours and fallback, Voice conversations, Transcription minutes, Qualification playbooks, Meeting booking, Google and Microsoft calendars, Visitor timeline, Identity, Customers and companies, Leads and pipeline, Export, Workflows, Widget home, Tickets, Dashboard, Usage and spend, Calendar integrations, Your brand, Two-step sign-in, Erasure and export, Documentation, Email support
- **Growth:** Human conversations, Customer profiles, AI agents, Data retention, Teams and routing, Roles, Website widget, Design studio, Match my site, Proactive experiences, Experience targeting, Experience results, Shared inbox, Live floor, Typing and presence, Saved replies and shortcuts, AI answers, Agent copilot, Agent policy, Website crawling, Documents, Knowledge gaps, Website chat, Rich replies, The full surface set, Adaptive stage, Handoff to a person, Handoff brief, Service hours and fallback, Voice conversations, Transcription minutes, Qualification playbooks, Lead scoring, Meeting booking, Google and Microsoft calendars, Visitor timeline, Identity, Intent and journey stage, Customers and companies, Leads and pipeline, Export, Workflows, Run history, Widget home, Tickets, Ticket workspace, Reply-time targets, Dashboard, Insights, Usage and spend, Calendar integrations, Your brand, Two-step sign-in, Erasure and export, Documentation, Email support, Priority support
- **Pro:** Human conversations, Customer profiles, AI agents, Data retention, Teams and routing, Roles, Website widget, Design studio, Match my site, Proactive experiences, Experience targeting, Experience results, Shared inbox, Live floor, Typing and presence, Saved replies and shortcuts, AI answers, Agent copilot, Agent policy, Website crawling, Documents, Knowledge gaps, Website chat, Rich replies, The full surface set, Adaptive stage, Handoff to a person, Handoff brief, Service hours and fallback, Voice conversations, Transcription minutes, Qualification playbooks, Lead scoring, Meeting booking, Google and Microsoft calendars, Visitor timeline, Identity, Intent and journey stage, Customers and companies, Leads and pipeline, Export, Workflows, Run history, Widget home, Tickets, Ticket workspace, Reply-time targets, Dashboard, Insights, Traces, Usage and spend, Calendar integrations, Public API, Webhooks, Your brand, Two-step sign-in, Erasure and export, Custom roles, Documentation, Email support, Priority support

## Beyond an allowance

Paid plans draw on prepaid credit past an allowance: packs of 10, 25, 100, never expiring. A visitor is never cut off mid-conversation. Nothing is charged per lead qualified or per meeting booked. Free runs its AI on the customer's own provider key.

## Comparison

### AI & automation

| Feature | Free | Starter | Growth | Pro | Enterprise |
| --- | --- | --- | --- | --- | --- |
| AI answers | Included | Included | Included | Included | Included |
| Agent copilot | — | Included | Included | Included | Included |
| Agent policy | — | Included | Included | Included | Included |
| Bring your own model key | Included | — | — | — | Included |
| Website crawling | Included | Included | Included | Included | Included |
| Documents | — | Included | Included | Included | Included |
| Knowledge gaps | — | — | Included | Included | Included |
| Workflows | — | Included | Included | Included | Included |
| Run history | — | — | Included | Included | Included |

### Channels

| Feature | Free | Starter | Growth | Pro | Enterprise |
| --- | --- | --- | --- | --- | --- |
| Website chat | Included | Included | Included | Included | Included |
| Voice conversations | — | Included | Included | Included | Included |
| AI voice minutes | – | 90 min | 250 min | 500 min | Custom |
| Transcription minutes | Your own AI key | 100 min | 300 min | 1,000 min | Custom |

### Customer experience

| Feature | Free | Starter | Growth | Pro | Enterprise |
| --- | --- | --- | --- | --- | --- |
| Websites | 1 | 2 | 10 | Unlimited | Unlimited |
| External AI credits | Your own AI key | 3,500 | 11,500 | 30,500 | Custom |
| Human conversations | Unlimited | Unlimited | Unlimited | Unlimited | Unlimited |
| Customer profiles | 1,000 | 10,000 | 100,000 | 500,000 | Custom |
| AI agents | 1 | 1 | 5 | Unlimited | Unlimited |
| Data retention | 30 | 90 | 365 | 730 | Custom |
| File storage | 100 MB | 2 GB | 10 GB | 30 GB | Custom |
| Website widget | Included | Included | Included | Included | Included |
| Design studio | Included | Included | Included | Included | Included |
| Match my site | — | Included | Included | Included | Included |
| Proactive experiences | — | Included | Included | Included | Included |
| Experience targeting | — | Included | Included | Included | Included |
| Experience results | — | — | Included | Included | Included |
| Shared inbox | Included | Included | Included | Included | Included |
| Live floor | — | Included | Included | Included | Included |
| Typing and presence | Included | Included | Included | Included | Included |
| Saved replies and shortcuts | — | Included | Included | Included | Included |
| Rich replies | Included | Included | Included | Included | Included |
| The full surface set | — | Included | Included | Included | Included |
| Adaptive stage | — | — | Included | Included | Included |
| Handoff to a person | Included | Included | Included | Included | Included |
| Handoff brief | — | Included | Included | Included | Included |
| Service hours and fallback | — | Included | Included | Included | Included |
| Widget home | Included | Included | Included | Included | Included |
| Your brand | Included | Included | Included | Included | Included |

### Support & ticketing

| Feature | Free | Starter | Growth | Pro | Enterprise |
| --- | --- | --- | --- | --- | --- |
| Tickets | — | Included | Included | Included | Included |
| Ticket workspace | — | — | Included | Included | Included |
| Reply-time targets | — | — | Included | Included | Included |

### Sales & qualification

| Feature | Free | Starter | Growth | Pro | Enterprise |
| --- | --- | --- | --- | --- | --- |
| Qualification playbooks | — | Included | Included | Included | Included |
| Lead scoring | — | — | Included | Included | Included |
| Meeting booking | — | Included | Included | Included | Included |
| Google and Microsoft calendars | — | Included | Included | Included | Included |
| Visitor timeline | — | Included | Included | Included | Included |
| Identity | — | Included | Included | Included | Included |
| Intent and journey stage | — | — | Included | Included | Included |
| Customers and companies | Included | Included | Included | Included | Included |
| Leads and pipeline | — | Included | Included | Included | Included |
| Export | — | Included | Included | Included | Included |

### Analytics

| Feature | Free | Starter | Growth | Pro | Enterprise |
| --- | --- | --- | --- | --- | --- |
| Dashboard | Included | Included | Included | Included | Included |
| Insights | — | — | Included | Included | Included |
| Traces | — | — | — | Included | Included |
| Usage and spend | — | Included | Included | Included | Included |

### Integrations

| Feature | Free | Starter | Growth | Pro | Enterprise |
| --- | --- | --- | --- | --- | --- |
| Calendar integrations | — | Included | Included | Included | Included |
| Public API | — | — | — | Included | Included |
| Webhooks | — | — | — | Included | Included |

### Security & administration

| Feature | Free | Starter | Growth | Pro | Enterprise |
| --- | --- | --- | --- | --- | --- |
| Team users | 1 | 3 | 10 | 25 | Custom |
| Teams and routing | — | Included | Included | Included | Included |
| Roles | Included | Included | Included | Included | Included |
| Two-step sign-in | Included | Included | Included | Included | Included |
| Erasure and export | — | Included | Included | Included | Included |
| Custom roles | — | — | — | Included | Included |
| Documentation | Included | Included | Included | Included | Included |
| Email support | — | Included | Included | Included | Included |
| Priority support | — | — | Included | Included | Included |

## Questions

### How are team users counted?

Everyone who signs in counts once, whatever their role: 1 on Free, 3 on Starter, 10 on Growth and 25 on Pro. Conversations your team answers are never counted, on any plan.

### What are external AI credits?

What the agent spends answering visitors and customers. A reply costs credits by the model that wrote it: 1 on Luna, 5 on Terra and 12 on Sol, more for an unusually heavy one. A conversation your own team answers uses none, on any plan, including the free one. Credits reset each month and do not roll over.

### What happens when we use more than the plan includes?

A visitor is never cut off mid-conversation because a counter ran out. On a paid plan, usage past the allowance draws on prepaid credit you top up in Settings → Billing: packs of $10, $25, $100, never expiring. It is never an open-ended bill, and nothing extra is charged unless you buy credit. Allowances start again when your plan renews.

### How does Free pay for its AI?

It runs on your own provider key — OpenAI, Anthropic, Google or Groq — connected in Settings → AI, so the AI on Free is billed by your provider, not by us. Human conversations are unlimited on Free, as on every plan.

### Which channels are included?

The website widget, on every plan, with voice conversations in the browser from Starter, counted in voice minutes.

### Are we charged per lead or per meeting booked?

No. Qualification, scoring, booking and handover are features of the plan. Charging per outcome would mean the product earns more the more work it makes for you, which is the wrong incentive to build on.

### Can we change plan or stop?

Yes, at any time. A change takes effect as soon as the payment is confirmed: the new plan's allowances replace the old ones for the rest of the period, what you have already used still counts, and the charge is prorated. If you cancel, your data stays yours: export is included from Starter.


---

# Enterprise

> Customer conversations, AI agents and your team together in one connected workspace.

- Canonical page: https://hispecular.com/enterprise
- Product: Specular, by SynconAI
- Updated: 2026-10-22

## What it gives a larger team

- One enquiry, three hands: the agent answers, a teammate takes over with a brief, and the customer record keeps what happened.
- One customer record across conversations, companies, leads, opportunities, meetings and tasks.
- Governance: agent policy sets subjects that always go to a person and things the agent may never promise; changes are published deliberately.
- Operational control: workflows with a run history; insights on what the agent answered, handed over and booked.
- Roles: Owner, Admin, Manager, Teammate and Viewer, with custom roles and field-level permissions on Pro and Enterprise.
- Your own AI provider key on Enterprise.

## Talk to sales

Write to contact@synconai.com. Enterprise is priced with you: AI allowances set to your volume.

## Not offered today

- Single sign-on (SAML or OIDC)
- SCIM user provisioning
- A browsable audit log in the workspace
- A choice of hosting region
- SOC 2, ISO 27001 or other third-party certification of Specular
- A contractual uptime commitment


---

# Security & Trust

> How Specular protects what customers tell a business, described as it works today.

- Canonical page: https://hispecular.com/enterprise/security
- Product: Specular, by SynconAI
- Updated: 2026-10-22

## Data protection

Your conversations, customer records and files belong to your workspace, and you can take them out or remove them.

- Conversations, people, companies, leads and files are stored against one workspace.
- Records can be exported as CSV by people whose role allows it.
- A person, visitor, conversation or the whole workspace can be deleted on request.
- Every conversation is kept in full, agent and teammates together, so you can see exactly what a customer was told.

**How deletion runs.** A deletion request names one subject: a person, a visitor, a conversation or the workspace. A background worker carries it out after a 24-hour grace period, so a mistaken request can be withdrawn the same day.

## Authentication and permissions

People sign in to their own accounts, can add a second step, and see only what their role allows.

- Two-step sign-in with an authenticator app, with backup codes stored encrypted.
- Sign-in, sign-up and password reset are rate limited for each IP address.
- Five roles in every workspace: Owner, Admin, Manager, Teammate and Viewer.
- Custom roles with field-level permissions on Pro and Enterprise.
- Permissions are checked on the server for every action, never by hiding a button.

**Sign-in limits.** In production, sign-in is limited to 8 attempts a minute, sign-up to 5 every five minutes and password reset to 3 every five minutes. The counters are kept in the database, so every server instance shares them.

## Workspace isolation

Each business is its own workspace, and one workspace cannot read another's conversations or records.

- Every query for workspace data is scoped to the workspace on the server.
- The workspace is taken from the signed-in session, not from anything the browser sends.
- Public API keys belong to one workspace and reach only that workspace's data.

## AI data handling

The agent answers from material you give it, records where each answer came from, and sends conversation text to a model provider to write its replies.

- Conversation text and the relevant passages from your material are sent to the AI model provider to produce each reply.
- On Free, and on Enterprise if you choose, the AI runs on your own provider key (OpenAI, Anthropic, Google or Groq), under your own contract with that provider.
- The source of each answer is recorded with it, so anything the agent said can be checked against what you published.
- Questions it could not answer are collected for your team rather than improvised.
- Agent policy sets subjects that always go to a person and things it may never promise.

## Encryption

Everything travels over TLS, and the secrets we hold for you are encrypted before they are stored.

- All traffic between visitors, your site, your team and the platform uses HTTPS.
- Browsers are told to use HTTPS only (HSTS, two years, including subdomains).
- Provider keys, calendar and mailbox tokens, channel credentials and webhook secrets are encrypted with AES-256-GCM before they reach the database.
- Public API keys are stored only as a hash; the full key is shown once, when it is created.

**What is and is not encrypted by the application.** Secrets are encrypted in the application with AES-256-GCM, a fresh random IV for each value, and an authentication tag checked on every read. Conversation text and customer records are protected by access control and transport encryption; we do not describe them as application-level encrypted, because they are not.

## Integrations and credentials

Connections use each provider's own sign-in, and what they hand back is stored encrypted.

- Google and Microsoft calendars and mailboxes connect through OAuth; we never see the account password.
- The OAuth state is signed and checked on return, so a connection cannot be completed by someone else's link.
- Outgoing webhooks are signed with HMAC-SHA256 over a timestamp and the body, so receivers can reject forged or replayed deliveries.

**Verifying a webhook.** Each delivery carries an x-syncon-signature header of the form t=<unix seconds>,v1=<hex HMAC-SHA256 of "t.body">. Reject a delivery whose timestamp is more than 300 seconds old or whose signature does not match, compared in constant time.

## Data retention and deletion

You choose how long closed conversations and uploads are kept; by default they are kept until you remove them.

- Closed conversations can be removed automatically after 30, 90, 180, 365 or 730 days.
- Uploaded files can be removed on their own schedule, from the same choices.
- Workspace exports are stored for seven days and then removed.
- Deletion requests run after a 24-hour grace period.

**Where this is set.** Settings → Data in the workspace. A policy is two numbers of days, or nothing, which means keep, and the worker applies it every hour. Anything outside 30 to 3,650 days is clamped into that range.

## Infrastructure and hosting

The application and its database run on servers SynconAI operates; files are kept in private object storage.

- The web application, background worker and PostgreSQL database run in containers on servers SynconAI operates.
- The database is backed up daily.
- Uploaded files and exports are stored in private Cloudflare R2 buckets with no public access.
- Email is sent through Resend; payments are taken by our payment provider.
- A hosting region cannot be chosen per workspace today.

## Auditability and governance

Consequential actions are recorded with who did them and when, and every answer can be traced to its source.

- Changes to connections, channels, knowledge, teammates and conversation ownership are written to an audit record with the person, the time, and the IP address and browser.
- Every handover to a person carries the reason and a brief.
- Changes to the agent are published deliberately rather than edited live in front of visitors.
- A browsable audit log in the workspace is not available today; records can be provided on request.

## Operational security

The site and the widget are hardened against the common web attacks, and input is checked at every boundary.

- A content security policy with a fresh nonce on every request.
- Pages cannot be framed by other sites, and browsers are told not to guess content types.
- The website widget renders inside a shadow root, so it cannot read or restyle your page.
- Everything that arrives from outside, from forms to the public API, is validated before it is used.

## Not offered today

- Single sign-on (SAML or OIDC)
- SCIM user provisioning
- A browsable audit log in the workspace
- A choice of hosting region
- SOC 2, ISO 27001 or other third-party certification of Specular
- A contractual uptime commitment

## Policies

- Privacy policy: https://hispecular.com/privacy
- Terms of service: https://hispecular.com/terms
- Refund policy: https://hispecular.com/refunds
- Security questions: contact@synconai.com


---

# Compare

> More than a chatbot. When Specular is the right choice, and when something else is.

- Canonical page: https://hispecular.com/compare
- Product: Specular, by SynconAI
- Updated: 2026-10-22

## The difference from a chatbot

A traditional chatbot follows flows someone drew, or answers without a source. Specular's agent answers from the business's own published material and keeps the source of each answer, qualifies while it helps, books on real availability, brings in a person with a brief, and writes what happened to a customer record and pipeline in the same workspace.

## Chatbot and flow builders

In their own words (read 22 October 2026):

- Landbot (https://landbot.io/): "Combine structured flows, AI flexibility, and branded UIs that convert."
- Manychat (https://manychat.com/): "Automate two-way, interactive conversations in Instagram Direct Messages, Facebook Messenger, and SMS."

- What they are good at: Designing exactly the questions a conversation asks, step by step, and running them on social messaging.
- Choose them when: Your customers mostly reach you on Instagram or Facebook Messenger, or you want every path drawn by hand. Specular does not connect to Instagram or Messenger.
- Where Specular differs: You do not draw flows. The agent answers from your own pages and documents, qualifies with the questions you set, books on your real calendar and hands over with a brief, and the customer record, opportunities and tasks are in the same workspace.

## Live chat and help desks with AI

In their own words (read 22 October 2026):

- tawk.to (https://www.tawk.to/): "One free inbox for all your customer channels. Live chat, ticketing, knowledge base, AI assist, and CRM."
- Zendesk AI agents (https://www.zendesk.com/service/ai/ai-agents/): "AI agents are autonomous systems that resolve customer service enquiries across channels like messaging, email and voice."
- Intercom with Fin (https://www.intercom.com/pricing): "Fin AI Agent on all plans for $0.99 per outcome."
- Tidio with Lyro (https://www.tidio.com/): "Lyro resolves questions instantly and turns the rest into sales, leads, and bookings you can attribute."

- What they are good at: Support teams working a shared inbox and tickets, with years of channel coverage and add-ons behind them.
- Choose them when: Support volume is the job and you need a mature help desk ecosystem, or you want free live chat and do not need an AI agent.
- Where Specular differs: Specular starts from the conversation that wins the customer as much as the one that supports them: qualification, scoring, booking and pipeline are built in, the agent has a character and a voice you choose, and AI pricing is a monthly allowance per plan rather than per resolution.

## AI SDR and inbound sales agents

In their own words (read 22 October 2026):

- Qualified (Piper) (https://www.qualified.com/pricing): "We'll build a plan that's perfectly customized for you and your pipeline needs."
- Warmly (https://www.warmly.ai/): "Identifies anonymous website visitors, enriches them with intent data, and engages buyers through AI chat and email."
- 1mind (https://www.1mind.com/): "AI sales agents that qualify inbound leads, deliver product demos, handle objections."

- What they are good at: Enterprise B2B pipeline: account-level intent signals, identifying companies behind anonymous traffic, and deep ties to an existing sales stack.
- Choose them when: You need anonymous company identification or third-party intent data, or a CRM-native agent inside an established Salesforce operation. Specular does not identify anonymous companies.
- Where Specular differs: Published prices and a free plan you can start today, one agent for sales and service together, its own customer record and pipeline rather than a dependency on another CRM, and a person brought in with a brief when it matters.

## CRM suites with chat

In their own words (read 22 October 2026):

- HubSpot (https://www.hubspot.com/products/crm): "HubSpot's free CRM unifies all your customer data on one platform, with AI that makes it easy to understand."

- What they are good at: Running marketing, sales and service from one customer database, with chat and AI agents as part of a much larger suite.
- Choose them when: Your teams already live in a CRM suite and want website chat inside it, configured by the people who run that suite.
- Where Specular differs: Specular leads with the conversation itself: an agent with a character and voice you choose, answers traced to your own material, a person brought in with a brief, and the customer record and pipeline built in rather than a suite to configure first.

## Enterprise AI customer service agents

In their own words (read 22 October 2026):

- Sierra (https://sierra.ai/): "Sierra helps businesses build better, more human customer experiences with AI."
- Decagon (https://decagon.ai/): "Build, optimize, and scale AI agents that treat every customer like the only one."
- Ada (https://www.ada.cx/): "AI customer service agents for enterprise CX."

- What they are good at: Large service operations with complex procedures, many languages and a dedicated team to build and tune the agent.
- Choose them when: You run contact-centre scale support and want a vendor-led build. Specular is self-serve first.
- Where Specular differs: A business can set Specular up itself in an afternoon, see the agent on its own website before signing up, and pay a published monthly price.

## Questions worth asking any vendor

### Can I see it on my own website before I pay?

Ours: yes. Type your address on the Experience page and the agent appears over your homepage, answering from what it says. No account.

### Are the prices published?

Ours: yes. Free on your own AI key, then $29 to $200 a month, with Enterprise priced with you. Every allowance is written on the pricing page.

### What happens when the AI should not answer?

Ours: it says it does not know rather than inventing, and hands the conversation to a person with a brief of who, why and what is still open.

### Where does the lead go?

Ours: onto a customer record in the same workspace, scored, with the conversation, the meeting and the opportunity attached. Webhooks and the API take it anywhere else.

### Which integrations are native, and which are not?

Ours are listed with their status, including the ones without a connector yet.

### Which certifications do you hold?

Ask everyone. We do not claim certifications we do not hold, and will tell you plainly what we have today.


---

# Try it on your site

> See how a Specular agent looks on a website, as any of the ten agents.

- Canonical page: https://hispecular.com/experience
- Product: Specular, by SynconAI
- Updated: 2026-10-22

## The sample website

A sample website in a browser or on a phone with the real widget on it, as any of the ten agents. It runs entirely in the visitor's browser with scripted answers; no AI and nothing is sent anywhere.

## Your own homepage

A visitor can enter their own address. Specular fetches that public homepage and up to three of its stylesheets to take its name, colours, type and text; private and internal addresses are refused. Questions asked there are sent with that text to the AI provider to write the answer, capped in size and cost. The page text and the questions are not stored; the hostname and some counts are kept in logs. No account is needed.

## The workspace demo

A read-only demo of the workspace with sample data runs entirely in the browser, with no sign-in and no backend: https://hispecular.com/demo.


---

# Privacy Policy

Last updated: 2026-10-29

## The short version

- Specular is a product of SynconAI Pty Ltd. We collect what is needed to run your account and your workspace, and nothing for advertising.
- What your workspace holds (your conversations, customer records and files) belongs to your business. We process it to run the service for you.
- If you sign in with Google, we receive your name, email address and profile picture, and use them only to sign you in and show who you are in your workspace.
- We do not sell personal information, we do not show ads, and we do not use your workspace content or Google data to train AI models.

## 1. Who we are

Specular is built and run by SynconAI Pty Ltd (“SynconAI”). In this policy, "we" and "us" mean SynconAI Pty Ltd.

This policy covers hispecular.com, the Specular workspace you sign in to, the chat and voice widget businesses place on their websites, and the public demo.

## 2. Our two roles

For your account, your sign-in and our own website, we decide how personal information is used.

For what a workspace holds (conversations with a business's website visitors, its customer records, its files and settings), the business that owns the workspace decides, and we process that information on its behalf to provide the service.

If you talked to an agent on a business's website, that business is responsible for your conversation. Contact them about it, or write to us and we will pass your request on.

## 3. What we collect

- Your account: your name, email address and password. The password is stored only as a hash. If you turn on two-step sign-in, the authenticator secret and backup codes are stored encrypted.
- Sign in with Google: your name, email address, profile picture and whether Google has verified the email address. See "Google user data" below.
- Your workspace: the conversations your agents and team have with visitors, people, companies, leads, notes, bookings, uploaded files, the material your agent learns from (website pages and documents you point it to), and your settings.
- Connected services: if you connect a calendar or mailbox, the access tokens the provider gives us, stored encrypted, and the calendar or email data the feature you turned on needs. If you add your own AI provider key, it is stored encrypted.
- Billing: your plan, payment status and invoices. Card details are handled by our payment provider and never reach us.
- Website visitors who talk to an agent: the messages, any details the visitor chooses to give (such as a name, email address or phone number), the page they are on, and an identifier kept in the browser's storage so the conversation can continue. In a voice call, what is said is sent to the AI provider so the agent can understand and answer, and the conversation is kept as text.
- Signed-in devices: the IP address and browser of each sign-in, used to keep sessions secure and listed for you in Settings → Security.
- Our own website: anonymous events, such as which page was viewed and which button was pressed, with no cookie and no identifier. Browsers that send Global Privacy Control or Do Not Track send nothing.

## 4. How we use it

- To provide Specular: sign you in, run your workspace, let your agents answer visitors and hand over to your team, and carry out the features you turn on.
- To keep the service secure: rate-limit sign-in attempts, detect abuse and investigate problems.
- To send the email the service needs: sign-in and password reset links, invitations, receipts, and reports you schedule.
- To take payment for paid plans and credit.
- To answer you when you contact us.
- To understand, in aggregate, how our own website is used, from the anonymous events above.

We do not sell personal information. We do not use it for advertising. We do not use your workspace content or Google user data to train AI models.

## 5. Google user data

When you choose "Continue with Google", Google shares your name, email address, profile picture and email verification status with us (the openid, email and profile scopes). We do not receive your Google password, and signing in does not give us access to your Gmail, Drive, Contacts, Calendar or any other Google data.

We use this information only to create your Specular account, sign you in, and show your name and picture to the people in your workspace. It is stored with your account and deleted when your account is deleted.

If you separately connect a Google calendar or mailbox to a workspace, you are asked for that access on Google's own screen, and the data is used only to provide the feature you connected, such as offering free times and booking meetings.

Specular's use and transfer to any other app of information received from Google APIs will adhere to the Google API Services User Data Policy, including the Limited Use requirements.

- We use Google user data only to provide and improve the features you can see and use.
- We do not transfer it to others except as needed to provide those features, to comply with the law, or as part of a merger or acquisition with notice to you.
- We do not use it for advertising, and we do not sell it.
- No person at SynconAI reads it unless you ask us to, it is needed for security or to comply with the law, or it has been aggregated and anonymised.

You can disconnect Google at any time in Settings → Security, and remove Specular's access from your Google account at myaccount.google.com/permissions.

## 6. Who we share it with

Inside your workspace, your teammates see what their role allows. Outside it, we use these services to run Specular, and each one receives only what its job needs:

- OpenAI writes the agent's replies and carries voice calls. On Free, and on Enterprise if you choose, the AI runs on your own provider key (OpenAI, Anthropic, Google or Groq) under your own contract with that provider.
- Cloudflare stores uploaded files and exports in private buckets.
- Resend sends the service's email.
- Our payment provider takes payment for paid plans and credit.
- Google and Microsoft, when you sign in with them or connect a calendar or mailbox.

We also disclose information when the law requires it, to protect people or the service from harm, or to a buyer if SynconAI or Specular is sold, in which case this policy continues to apply to it.

## 7. Where it is stored

The application and its database run on servers SynconAI operates, and the database is backed up daily. Files and exports are kept in private Cloudflare R2 storage. Your information may be processed in countries other than yours, including the United States, and we protect it the same way wherever it is.

## 8. How long we keep it

- Your account and workspace are kept while they are open.
- A workspace can remove closed conversations and uploaded files automatically after a period it chooses, in Settings → Data. Otherwise they are kept until removed.
- Workspace exports are deleted after seven days.
- A request to delete a person, a visitor, a conversation or a whole workspace is carried out after a 24-hour grace period, so a mistaken request can be withdrawn.
- Deleted information can remain in our daily database backups until those backups are replaced.

## 9. How we protect it

Everything travels over HTTPS. Passwords are hashed; access tokens, provider keys and other secrets are encrypted before they are stored; every request for workspace data is checked against the signed-in person's workspace and role. The full detail is on our Security & Trust page at hispecular.com/enterprise/security.

## 10. Your choices and rights

- See and correct your account details in Settings.
- Export your workspace's records as CSV, on a paid plan.
- Ask us to delete your account, or ask your workspace's owner to delete the workspace.
- Disconnect Google or Microsoft sign-in in Settings → Security.
- Ask us for a copy of the personal information we hold about you, or object to how we use it.

Write to contact@synconai.com and a person will answer. Depending on where you live, you may also complain to your data protection authority.

## 11. Cookies and browser storage

Specular sets one kind of cookie: the sign-in session cookie that keeps you signed in. The widget on a business's website keeps the conversation in the browser's storage so it can continue on the next page. We use no advertising cookies and no third-party trackers, and our own website analytics use no cookie at all.

## 12. Children

Specular is a service for businesses and is not meant for anyone under 16. We do not knowingly collect information from children. If you believe a child has given us information, write to us and we will delete it.

## 13. Changes to this policy

When we change this policy we update the date at the top. If a change affects how we use your information in a meaningful way, we tell account holders by email or in the product before it takes effect.

## 14. Contact

SynconAI Pty Ltd.

Email: contact@synconai.com


---

# Terms of Service

Last updated: 2026-10-29

## The short version

- These terms are the agreement between you and SynconAI Pty Ltd for using Specular.
- Your content and your customers' data stay yours. You are responsible for how you use the agents and for telling your visitors about them.
- AI agents can be wrong. Set them up with care and check what matters.
- Paid plans renew until you cancel. Payments are taken by our payment provider, and refunds are in our Refund Policy.

## 1. The agreement

These terms are an agreement between you and SynconAI Pty Ltd (“SynconAI”) for the use of Specular: hispecular.com, the Specular workspace, its widget, API and integrations. If you use Specular for an organisation, you accept these terms for that organisation and confirm you may do so.

If your organisation has signed a separate written agreement with SynconAI for Specular, that agreement takes precedence where the two differ.

Our Privacy Policy, at hispecular.com/privacy, explains how we handle personal information and forms part of these terms.

## 2. Accounts and workspaces

- You must be at least 16 and able to enter into a contract to create an account.
- Give accurate details, keep your password and sign-in methods safe, and tell us if you think someone else has used your account.
- A workspace belongs to the organisation that created it. Its Owner controls who is in it and what each role can do, and is responsible for the activity in it.
- You are responsible for what happens under your account.

## 3. Your content and your customers' data

You keep all rights to what you put into Specular and what your workspace collects: conversations, customer records, files and the material your agents learn from. You give us permission to host, copy, process and display it only as needed to provide Specular to you and as our Privacy Policy describes.

You are responsible for having the right to collect and use the information your workspace gathers from your visitors and customers, for telling them how you use it (including that an AI agent may answer them), and for meeting the laws that apply to you.

You can export your records and delete your workspace as described in the product and our Privacy Policy.

## 4. AI agents

Agents write their answers with AI from the material and instructions you give them. AI can be wrong, incomplete or out of date. You decide what your agents may say and do, which subjects always go to a person, and what they may never promise, and you remain responsible for what your agents tell your customers.

Do not rely on an agent's answer for medical, legal, financial or other professional advice, and check anything important before you act on it.

## 5. Acceptable use

Do not use Specular to:

- break the law, or help someone else break it;
- send spam or unsolicited messages, or contact people who have not agreed to hear from you;
- harass, deceive or impersonate anyone, or pretend an agent is a person when asked;
- collect payment card numbers, passwords or other information you are not allowed to hold;
- upload malware, or try to break into, overload or disrupt Specular or anyone's use of it;
- copy, resell or reverse engineer the service, or get around its limits or security.

We may remove content or suspend access that breaks these rules.

## 6. Plans, billing and credit

- Plans, prices and allowances are listed at hispecular.com/pricing. Paid plans are billed monthly or annually, in advance, and renew until you cancel.
- Payments, receipts and invoices are handled by our payment provider, and its checkout terms apply to the payment.
- You can change or cancel your plan at any time. A change takes effect once the payment is confirmed and is prorated; a cancellation stops the next renewal.
- Usage past your plan's allowance draws on prepaid credit you choose to buy. Credit is sold at face value and does not expire. Nothing extra is charged unless you buy credit.
- On Free, the AI runs on your own provider key, and that provider's charges and terms apply to it.
- Prices do not include taxes unless the checkout says so.

Refunds are covered by our Refund Policy, at hispecular.com/refunds. If you think a charge is wrong, write to contact@synconai.com.

## 7. Other services

Specular connects to services run by others, such as Google, Microsoft, AI model providers and the tools you link through webhooks and the API. Their own terms apply to your use of them, and we are not responsible for them.

## 8. Availability and changes

We work to keep Specular running and secure, but we do not promise it will always be available or free of errors, and we do not offer a service level agreement unless one is part of a separate written agreement. We may change, add or remove features. If we remove something you pay for in a way that matters, we will tell you in advance where we reasonably can.

## 9. Suspension and ending

You can stop using Specular and close your workspace at any time. We may suspend or close an account or workspace that breaks these terms, puts others at risk or is not paid for, and we will tell you why unless the law or security prevents it. Export what you need before you close a workspace; after that it is deleted as our Privacy Policy describes.

## 10. Our service and your feedback

Specular, its software, design, characters and name belong to SynconAI. These terms give you the right to use the service, not ownership of it. If you send us ideas or feedback, we may use them without owing you anything.

## 11. Disclaimers

Except as these terms say, Specular is provided "as is" and "as available", without warranties of any kind, to the extent the law allows.

## 12. Limitation of liability

To the extent the law allows, SynconAI is not liable for indirect or consequential loss, or for loss of profit, revenue, business, goodwill or data. SynconAI's total liability arising from Specular is limited to the amount you paid for it in the twelve months before the claim, or US$100 if you paid nothing. Nothing in these terms limits liability that cannot be limited by law.

## 13. Your responsibility to us

If someone brings a claim against SynconAI because of your content, your use of Specular in breach of these terms, or your breach of the law, you will cover the reasonable costs of that claim.

## 14. Governing law

These terms are governed by the laws of the State of Delaware, United States, and the courts of that state have non-exclusive jurisdiction.

## 15. Changes to these terms

When we change these terms we update the date at the top. If a change matters, we tell account holders by email or in the product before it takes effect. Continuing to use Specular after that means you accept the change.

## 16. Contact

SynconAI Pty Ltd.

Email: contact@synconai.com


---

# Refund Policy

Last updated: 2026-10-29

## The short version

- Ask for a refund within 14 days of a payment and you get that payment back in full, taxes included.
- Credit you bought and have not used is refunded in the same 14 days. Credit already used is not.
- You can cancel at any time. A plan runs to the end of the period you paid for, and nothing more is charged.
- Payments and refunds are handled by our payment provider, and go back to the card or account you paid with.

## 1. Refunds within 14 days

- This covers every payment for Specular: the first payment for a plan, a renewal, and a credit pack.
- Ask within 14 days of the payment date and the whole payment is refunded, including the tax charged on it.
- A refund does not cancel your plan by itself. If you also want the plan to stop, cancel it from Billing, or tell us and we will.

## 2. Prepaid credit

- Credit that has not been used is refunded in full within the 14 days. When the refund is paid, that credit comes out of your balance.
- Credit that has already been used is not refundable. If a refund covers credit you have spent, only the unused part is returned.
- Credit does not expire, so after 14 days there is no refund for it, but it stays in your balance until you use it.

## 3. After 14 days

- You can cancel a plan at any time. It stays on until the end of the period you paid for, then your workspace moves to Free, and nothing more is charged.
- We do not refund the part of a period that has already been paid for once the 14 days have passed.
- If we charged you in error, for example twice or the wrong amount, we refund the difference at any time.

## 4. How to ask

- Write to contact@synconai.com from the email address on your account. Say which workspace and which payment, and we will arrange it.
- You can also ask our payment provider directly, from the receipt it emailed you.
- The refund is paid to the original payment method. How quickly it shows depends on your bank or card provider, usually within a few working days.

Please write to us before disputing a payment with your bank. A dispute that is upheld also takes the credit bought with that payment out of your balance.

## 5. Who takes the payment

Payments, receipts, invoices, taxes and refunds are handled by our payment provider. Its checkout terms apply to the payment, and this policy sits alongside them.

Specular is run by SynconAI Pty Ltd. Questions about this policy go to contact@synconai.com.


---

# Specular documentation

> Task guides for setting up and running Specular, in the words the workspace uses.

- Canonical page: https://hispecular.com/docs
- Updated: 2026-10-22

## Getting started

- [Introduction](https://hispecular.com/docs/introduction.md): What Specular does, how a workspace is laid out, and where to begin.
- [Create a workspace](https://hispecular.com/docs/create-a-workspace.md): Sign up, choose a plan, and follow the getting-started checklist on Home.
- [Configure your first agent](https://hispecular.com/docs/first-agent.md): Add your website, choose an agent, teach it your content and publish it.
- [Test and go live](https://hispecular.com/docs/go-live.md): Send it a test message, add the script to your site, and turn the site on.

## Agents

- [Agent personalities](https://hispecular.com/docs/agent-personalities.md): Choose one of the ten agents and adjust how it talks and moves.
- [Voice settings](https://hispecular.com/docs/voice-settings.md): Pick the voice your agent speaks with, and its pace, energy and answer length.
- [Behaviour and instructions](https://hispecular.com/docs/behaviour.md): Tell the agent who it is, what your business offers, and what it must never say.
- [Knowledge](https://hispecular.com/docs/knowledge.md): Give the agent your website pages and documents to answer from, and fill its gaps.

## Customer experience

- [Website widget](https://hispecular.com/docs/website-widget.md): The script tag, its attributes, placing it inside a page, and the JavaScript API.
- [Inbox](https://hispecular.com/docs/inbox.md): Work conversations with your team: assign, prioritise, tag and note.
- [Human handoff](https://hispecular.com/docs/human-handoff.md): When the agent brings in a person, how your team joins, and how to hand back.
- [Bookings](https://hispecular.com/docs/bookings.md): Connect a calendar so the agent offers real times and books meetings.

## Workspace

- [Customers and leads](https://hispecular.com/docs/customers-and-leads.md): People, companies and leads: where they come from and how they are qualified.
- [Opportunities](https://hispecular.com/docs/opportunities.md): Track deals through your pipeline stages, from first conversation to won or lost.
- [Meetings](https://hispecular.com/docs/meetings.md): Meetings booked by the agent or your team, and what comes with them.
- [Automations](https://hispecular.com/docs/automations.md): Workflows that act on what happens: assign, tag, prioritise, escalate.
- [Reporting](https://hispecular.com/docs/reports.md): The Home dashboard and Insights: what the agent answered, handed over and booked.

## Administration

- [Team access](https://hispecular.com/docs/team-access.md): Invite teammates, choose their role, and set custom permissions.
- [Integrations](https://hispecular.com/docs/integrations.md): Calendars, mailboxes and your own AI provider, connected from Settings.
- [Billing and usage](https://hispecular.com/docs/billing-and-usage.md): Your plan, what is counted, prepaid credit and invoices.
- [Security and privacy](https://hispecular.com/docs/security-and-privacy.md): Two-step sign-in, retention, and erasing a person's data.

## Developers

- [Public API](https://hispecular.com/docs/public-api.md): Read and write customers, deals, tickets, meetings and the live chat inbox over REST, with an OpenAPI document to build against.
- [Authentication and scopes](https://hispecular.com/docs/api-authentication.md): Create an API key, choose its scopes and role, act as a teammate, and keep keys safe.
- [Requests and responses](https://hispecular.com/docs/api-conventions.md): Pages, filters, syncing, error codes, retries with idempotency keys, and rate limits.
- [Webhooks](https://hispecular.com/docs/webhooks.md): Receive signed events when records change, subscribe from code, and verify each delivery.
- [Recipes](https://hispecular.com/docs/api-recipes.md): Common jobs, start to finish: a customer and a deal, reacting to a win, tickets, meetings, syncing.
- [The live chat inbox](https://hispecular.com/docs/api-live-chat.md): Read conversations, take them over from the AI agent, reply, hand back, and organise them through the API.
- [MCP server for AI agents](https://hispecular.com/docs/api-mcp.md): Let Claude, ChatGPT, Cursor or your own agent work the workspace, limited to what its key allows.
- [Connect an automation tool](https://hispecular.com/docs/api-automation-tools.md): Use Zapier, Make, n8n, Pabbly or similar with HTTP and webhook steps, or build Specular into your own product.


---

# Introduction

> What Specular does, how a workspace is laid out, and where to begin.

- Canonical page: https://hispecular.com/docs/introduction
- Section: Getting started
- Updated: 2026-10-22

Specular puts an AI agent on your website that answers from your own pages and documents, qualifies the people who are worth following up, books meetings on your real calendar and brings in a teammate when a person is needed. Everything it learns lands in the same workspace your team works from.

## What it does

- Answers visitors from your website and documents, and keeps the source of each answer.
- Asks the questions you choose while it helps, and turns the answers into a lead.
- Offers real times from a connected calendar and books the meeting.
- Hands a conversation to your team with a brief, when it should not answer alone.

## How the workspace is laid out

Home shows what is happening and what needs you. Inbox holds every conversation. Websites is where each site's agent is set up: its behaviour, its character, its design, how visitors meet it, a place to test it, and the script to install. Customers, companies, leads, opportunities and meetings are the records the conversations fill in. Settings holds your team, billing, connections and data controls.

## Where to start

Create a workspace, then work down the getting-started checklist on Home. The next three guides follow it in order.

1. [Create a workspace](https://hispecular.com/docs/create-a-workspace)
2. [Configure your first agent](https://hispecular.com/docs/first-agent)
3. [Test and go live](https://hispecular.com/docs/go-live)


---

# Create a workspace

> Sign up, choose a plan, and follow the getting-started checklist on Home.

- Canonical page: https://hispecular.com/docs/create-a-workspace
- Section: Getting started
- Updated: 2026-10-22

A workspace holds one business: its websites, agents, conversations, records and team.

## Sign up

1. Go to [hispecular.com/signup](https://hispecular.com/signup) and create your account.
2. Name your workspace. You can rename it later in Settings.
3. You land on Home, with the getting-started checklist at the top.

## The getting-started checklist

The checklist on Home tracks what stands between you and a live agent. Each step opens the screen where it is done, and ticks itself when the work is finished.

- **Add a website.** Register the domain the agent will run on.
- **Connect an AI provider.** Only on plans that run on your own AI key. Until one is connected the agent cannot answer anybody.
- **Publish the agent.** Name it and review its company guide.
- **Teach it your content**, **Arrange what visitors meet**, **Decide what to ask** and **Connect a calendar** are optional, and make it better at its job.
- **Send it a test message**, **Install the snippet** and **Go live** finish the job.

## Plans and your AI key

Free runs the agent on your own AI provider key, connected in Settings → AI. Paid plans include a monthly AI allowance, so no key is needed. The [pricing page](https://hispecular.com/pricing) lists every allowance.

> **On Free, connect a key first** Without a key the agent cannot reply on Free, so the checklist puts this step second and holds the rest until it is done.


---

# Configure your first agent

> Add your website, choose an agent, teach it your content and publish it.

- Canonical page: https://hispecular.com/docs/first-agent
- Section: Getting started
- Updated: 2026-10-22

This walks through setting up one website's agent from nothing to published. It takes about fifteen minutes if your website already says what you do.

## Add your website

1. Open Websites and add your site's domain.
2. The site opens on its Overview, with tabs for everything below.

## Choose the agent

Open the Agent tab and choose one of the ten agents. Ray is chosen to start with. Each has its own face, voice and way of talking; every one can do the same work. See [Agent personalities](https://hispecular.com/docs/agent-personalities).

## Teach it your content

Open Knowledge and add your website so its pages are read, and upload the documents customers ask about: price lists, policies, guides. The agent answers from what is published here. See [Knowledge](https://hispecular.com/docs/knowledge).

## Set its behaviour

Open Behaviour. Give it a name, role, tone and greeting, describe your business in the company guide, and write down the claims it must never make. Once your knowledge has enough text in it, each field can be drafted from it for you to edit. See [Behaviour and instructions](https://hispecular.com/docs/behaviour).

## Publish

Press **Publish agent** on Behaviour. Visitors only ever meet the published version, so you can change things and publish again when you are happy. Then [test it and go live](https://hispecular.com/docs/go-live).


---

# Test and go live

> Send it a test message, add the script to your site, and turn the site on.

- Canonical page: https://hispecular.com/docs/go-live
- Section: Getting started
- Updated: 2026-10-22

Three steps between a published agent and visitors talking to it.

## Send it a test message

Open the site's Test tab and ask what a customer would ask. The message goes through the same path a visitor's does, so the answer you read is the one they would get. The checklist counts one answered test as a step before going live.

## Install the script

Open the Deploy tab and copy the script tag. It carries your site's own id. Paste it before the closing `</body>` tag on every page, or in your site builder's custom code setting.

```html
<script src="https://hispecular.com/widget/v1/loader.js" data-site-id="YOUR_SITE_ID" async></script>
```

> **Use the tag from your Deploy tab** The example above shows the shape. Your Deploy tab has the exact tag with your site id in it, and the checklist ticks **Install the snippet** once the widget has loaded from your domain.

## Go live

With the agent published and tested, enable the site from its Overview. Visitors on your registered domain can now open the agent. More on the script and what it accepts in [Website widget](https://hispecular.com/docs/website-widget).


---

# Agent personalities

> Choose one of the ten agents and adjust how it talks and moves.

- Canonical page: https://hispecular.com/docs/agent-personalities
- Section: Agents
- Updated: 2026-10-22

Every website's agent is one of ten characters. They differ in how they look, sound and talk, not in what they can do: each answers from your knowledge, qualifies, books and hands over the same way.

## Choose an agent

1. Open Websites, choose the site, then the Agent tab.
2. Choose a character. Its face, voice and personality settings come with it.
3. Press **Customise** to adjust it. A customised character is marked as such, so you can always see which settings are your own.

## Personality and behaviour

Under **Personality & behaviour** you set the communication style, formality, friendliness, conversational warmth, initiative, humour, answer length in chat, and how often it asks questions. These shape how it says things; what it says comes from your [knowledge](https://hispecular.com/docs/knowledge) and [behaviour](https://hispecular.com/docs/behaviour).

## Avatar and motion

Under **Avatar & motion** you can slow it down or liven it up: animation pace, movement energy, idle movement, floating speed, bounce, tilt, and how it moves while speaking, listening and thinking. Under **Advanced** you can use your own image instead of the character.

> **Try it before you save** The live preview beside the settings shows every change as you make it, including each state: idle, listening, thinking and speaking.


---

# Voice settings

> Pick the voice your agent speaks with, and its pace, energy and answer length.

- Canonical page: https://hispecular.com/docs/voice-settings
- Section: Agents
- Updated: 2026-10-22

When visitors talk to the agent by voice, these settings decide how it sounds.

## Choose a voice

Open the site's Agent tab, press **Customise**, and find **Voice**. Each character starts with a voice that suits it; you can choose another from the list and press **Hear the voice** to compare.

## Pace, energy and length

**Speaking pace**, **Energy** and **Answer length on calls** shape what the agent says on a call. They are not speed or volume dials: a brisk, brief agent says less, sooner.

## Voice minutes

Voice conversations are included from Starter and count against the plan's monthly **AI voice minutes**. Minutes past the allowance draw on prepaid credit. See [Billing and usage](https://hispecular.com/docs/billing-and-usage).


---

# Behaviour and instructions

> Tell the agent who it is, what your business offers, and what it must never say.

- Canonical page: https://hispecular.com/docs/behaviour
- Section: Agents
- Updated: 2026-10-22

Behaviour is the agent's brief: who it is, what your business does, how it qualifies and hands over, and what it must never say. Visitors meet the published version.

## What you can set

- **Identity:** agent name, role, tone and greeting.
- **Your business:** the company guide, products and services, and who you sell to.
- **Conversation:** discovery guidance, the questions that tell a lead from a browser.
- **Guardrails:** claims the agent must not make, and the topics and boundaries it keeps to.

## Draft it from your knowledge

Once your [knowledge](https://hispecular.com/docs/knowledge) holds enough published text, each field offers a draft written only from it. Read and edit every draft; nothing is published until you press **Publish agent**.

## Workspace agent policy

In Settings, **Agent policy** sets rules for every agent in the workspace. They sit above each agent's own instructions, and changes are versioned and audited. **Fallback & escalation** sets when the agent asks a clarifying question, declines to state something it cannot support, or hands over to a person.

> **Guardrails are not a substitute for knowledge** An agent with no sources can only reason from the company guide. Give it your pages and documents so it can answer, and use guardrails for what it must not say.


---

# Knowledge

> Give the agent your website pages and documents to answer from, and fill its gaps.

- Canonical page: https://hispecular.com/docs/knowledge
- Section: Agents
- Updated: 2026-10-22

Knowledge is the material your agents answer from. It is shared by every agent in the workspace, and each answer keeps the source it came from.

## Add sources

1. Open Knowledge.
2. Add your website to have its pages read, choosing which pages to include.
3. Upload documents: PDF, Word (.docx), plain text, Markdown or CSV.
4. Publish what the agent should use. Unpublished sources are never quoted.

## How it answers from them

For each question the agent finds the relevant passages in your published sources and answers from them. In the inbox, an answer shows which source it came from, so anyone on your team can check what was said against what you published.

## Fill the gaps

From Growth up, **Knowledge gaps** on Home lists what visitors asked that the agent could not answer, so you know which page or document to add next.

> **Errors are listed, not hidden** If a page cannot be read or a file cannot be processed, Knowledge says so, and you can fix or disable the source before you publish.


---

# Website widget

> The script tag, its attributes, placing it inside a page, and the JavaScript API.

- Canonical page: https://hispecular.com/docs/website-widget
- Section: Customer experience
- Updated: 2026-10-22

The widget is one script. It loads your site's configuration, then the widget itself, and renders inside a shadow root, so your page's styles never reach it and its styles never reach your page.

## The script tag

Copy the tag from the site's Deploy tab and paste it before the closing `</body>` tag on every page where the agent should appear.

```html
<script src="https://hispecular.com/widget/v1/loader.js" data-site-id="YOUR_SITE_ID" async></script>
```

## Attributes

| Attribute | Required | What it does |
| --- | --- | --- |
| `data-site-id` | Yes | Which site the visitor is on. Your Deploy tab fills it in. |
| `data-embed` | No | A CSS selector. The widget renders inside that element instead of floating over the page. |
| `data-api-base` | No | Where the platform lives. Leave it out; it defaults to where the script loaded from. |

## Place it inside a page

To put the conversation in your layout, for example on a contact page, add an element and name it in `data-embed`. You can also choose the embedded layout for the site in Design.

```html
<div id="ask-us"></div>
<script src="https://hispecular.com/widget/v1/loader.js" data-site-id="YOUR_SITE_ID" data-embed="#ask-us" async></script>
```

## JavaScript API

Once the script is on the page, `window.SynconLead` lets your own code open and close the widget, ask on the visitor's behalf, or pass on who they are. Calls made before the widget has loaded are queued and run when it is ready.

```javascript
// Open or close the panel
window.SynconLead.open();
window.SynconLead.close();

// Open the panel and send a message as the visitor
window.SynconLead.ask("Do you deliver on Saturdays?");

// Tell the widget who is signed in on your site
window.SynconLead.identify({ email: "customer@example.com" });
```

> **identify is not proof of identity** `identify` fills in the visitor's email so they are not asked for it again. It is not verified, so do not use it to grant access to anything.


---

# Inbox

> Work conversations with your team: assign, prioritise, tag and note.

- Canonical page: https://hispecular.com/docs/inbox
- Section: Customer experience
- Updated: 2026-10-22

Inbox holds every conversation, whether the agent is handling it or a person is.

## The queue

Conversations are listed with who they are from, where they started and their state. Those waiting for a person are marked, so nobody misses a handover.

## Working a conversation

- Assign it to a teammate, or transfer it to another.
- Set its priority and add tags.
- Leave internal notes. Visitors never see them.
- See what the agent understood, and the source behind each of its answers.

## From conversation to record

From a conversation you can create the customer record it belongs to, and identity details a visitor gives are kept with them. See [Customers and leads](https://hispecular.com/docs/customers-and-leads).


---

# Human handoff

> When the agent brings in a person, how your team joins, and how to hand back.

- Canonical page: https://hispecular.com/docs/human-handoff
- Section: Customer experience
- Updated: 2026-10-22

The agent brings in a person when a conversation needs one, and your team can step in at any time.

## When it hands over

It hands over when a visitor asks for a person, when a subject is one your [agent policy](https://hispecular.com/docs/behaviour) says always goes to a person, or when it cannot answer well enough. The handover carries a brief: who is asking, why, and what is still open.

## Join a conversation

1. Open the conversation in Inbox.
2. Press **Join conversation**. Joining opens the composer and stops the agent replying.
3. Reply as yourself. The visitor sees that a person has joined.

## Hand back or transfer

When you are done, press **Hand back to agent** and it carries on from where you left off. To pass the conversation to someone else instead, transfer it.

> **Outside your hours** Set your team's hours in Settings → Hours. Outside them, the agent says when a person will be back instead of promising one now.


---

# Bookings

> Connect a calendar so the agent offers real times and books meetings.

- Canonical page: https://hispecular.com/docs/bookings
- Section: Customer experience
- Updated: 2026-10-22

With a calendar connected, the agent offers times your team is really free and books the meeting into their calendar. Without one it can only take a message.

## Connect a calendar

1. Open Calendar.
2. Connect Google Calendar or Microsoft 365 and approve access.
3. Each teammate who takes meetings connects their own calendar.

## Meeting types

Create a **New meeting type** for each kind of meeting you offer, such as a demo or a consultation: its length and location, the gaps around it, the notice it needs and how far ahead it can be booked.

## What visitors see

When a visitor wants to meet, the agent shows the available times in the conversation. Once they choose, the meeting is written to the calendar and appears in [Meetings](https://hispecular.com/docs/meetings).


---

# Customers and leads

> People, companies and leads: where they come from and how they are qualified.

- Canonical page: https://hispecular.com/docs/customers-and-leads
- Section: Workspace
- Updated: 2026-10-22

Every conversation belongs to a person. What they say, the company they work for, and whether they are worth following up all land on records your team shares.

## Customers and companies

Customers and Companies hold the people and organisations you deal with. A customer's record shows their conversations, opportunities, meetings and tasks together, and what the agent remembered from earlier conversations.

## Leads

Leads lists the people the agent or your team has judged worth following up, with an owner. Assignment rules decide who that owner is, and say why.

## Qualification and scoring

1. In Qualify, build a playbook: the few questions that separate a lead from a browser. The agent asks them while it helps, and publishes nothing until you do.
2. From Growth up, Score holds the scoring model that turns what was said into a score.

> **Routing never moves an owned lead** Once a lead has an owner, a new conversation from the same person does not reassign it.


---

# Opportunities

> Track deals through your pipeline stages, from first conversation to won or lost.

- Canonical page: https://hispecular.com/docs/opportunities
- Section: Workspace
- Updated: 2026-10-22

Opportunities tracks deals: their value, expected close date, owner and stage.

## The pipeline

Opportunities move through the stages of a pipeline. Pipelines and their stages are set in Settings → CRM → Pipelines, along with the reasons a deal can be won or lost.

## Moving through stages

Move an opportunity on its record or on the board. Won and lost stages close it, and the reason you give is kept with it for [reporting](https://hispecular.com/docs/reports).


---

# Meetings

> Meetings booked by the agent or your team, and what comes with them.

- Canonical page: https://hispecular.com/docs/meetings
- Section: Workspace
- Updated: 2026-10-22

Meetings lists meetings booked by the agent or your team, from connected calendars. See [Bookings](https://hispecular.com/docs/bookings) to connect one.

## Your meetings

Each meeting links to the person, their company and the opportunity it belongs to, so whoever takes it can see how it came about.

## Before the meeting

**Preparation** gathers what you need before a meeting: who you are meeting, what they asked, and what is still open.


---

# Automations

> Workflows that act on what happens: assign, tag, prioritise, escalate.

- Canonical page: https://hispecular.com/docs/automations
- Section: Workspace
- Updated: 2026-10-22

Automations holds your workflows: when something happens, do something about it.

## Workflows

A workflow has a trigger, optional conditions, and the actions to take, such as assigning a conversation, tagging it, setting its priority or escalating it. Workflows are included from Starter.

## Run history

From Growth up, every run is recorded with each step and where it stopped, so you can see what was done on your behalf and why.


---

# Reporting

> The Home dashboard and Insights: what the agent answered, handed over and booked.

- Canonical page: https://hispecular.com/docs/reports
- Section: Workspace
- Updated: 2026-10-22

Two places tell you how it is going: Home, for today, and Insights, for the trend.

## Home

Home shows who is on your site now, who is waiting for a person, the getting-started checklist until you are live, knowledge gaps and answer quality.

## Insights

Insights covers what the agent answered, handed over and booked over time, with reports and dashboards you can build and share. Insights is part of Growth and up.


---

# Team access

> Invite teammates, choose their role, and set custom permissions.

- Canonical page: https://hispecular.com/docs/team-access
- Section: Administration
- Updated: 2026-10-22

Everyone who signs in to the workspace is a team user, whatever their role.

## Invite teammates

1. Open Settings → Members.
2. Under **Invite**, enter their email and choose a role.
3. They get an email to join. Each counts once towards your plan's team users.

## Roles

| Role | For |
| --- | --- |
| Owner | Everything, including billing. |
| Admin | Runs the workspace: settings, members, websites and agents, connections and every record. |
| Manager | Leads the team: websites and agents, every conversation and record, CRM setup, and inviting teammates. |
| Teammate | Works the inbox and customer records; sees websites and agents without changing them. |
| Viewer | Reads conversations, records, websites and reports without changing anything. |

## Custom permissions

On Pro and Enterprise, Settings → Permissions lets you build permission sets, down to which fields a role can see or change, and assign them to people.


---

# Integrations

> Calendars, mailboxes and your own AI provider, connected from Settings.

- Canonical page: https://hispecular.com/docs/integrations
- Section: Administration
- Updated: 2026-10-22

Connections are made from Settings → Integrations and Calendar. Each uses the provider's own sign-in; Specular never sees the account password.

## Calendars and mail

- **Google Calendar** and **Microsoft 365** calendars, so the agent offers real times and books them. See [Bookings](https://hispecular.com/docs/bookings).
- **Gmail** and **Microsoft 365 mail**, so customer email lands on the customer's record and replies go from your own address.

## Your own AI provider

On Free, and on Enterprise if you choose, the agent runs on your own provider key: OpenAI, Anthropic, Google or Groq. Add it in Settings → AI under **Your AI provider**. The key is encrypted before it is stored.

## Everything else

For tools without a built-in connection, use the [public API](https://hispecular.com/docs/public-api) and [webhooks](https://hispecular.com/docs/webhooks), both on Pro. The [integrations page](https://hispecular.com/integrations) lists every connection and its status.


---

# Billing and usage

> Your plan, what is counted, prepaid credit and invoices.

- Canonical page: https://hispecular.com/docs/billing-and-usage
- Section: Administration
- Updated: 2026-10-22

Settings → Billing shows your plan, what you have used this month, your credit balance, and your invoices.

## Your plan

Change plan from Billing at any time. A change takes effect once payment is confirmed, the new allowances apply for the rest of the period, and the charge is prorated.

## What is counted

AI credits for the agent's replies to visitors, AI credits for Ray, voice minutes, transcription minutes, file storage, websites and team users each have a monthly allowance on your plan. Conversations your own team answers are never counted. The [pricing page](https://hispecular.com/pricing#what-is-counted) lists every number.

## Prepaid credit

Past an allowance, paid plans draw on prepaid credit. Buy packs of $10, $25 or $100 in Billing; they do not expire. A visitor is never cut off mid-conversation because a counter ran out.

> **Free runs on your own key** On Free, AI usage is billed by your AI provider, not by us.


---

# Security and privacy

> Two-step sign-in, retention, and erasing a person's data.

- Canonical page: https://hispecular.com/docs/security-and-privacy
- Section: Administration
- Updated: 2026-10-22

The controls you manage yourself. For how the platform protects your data, see [Security & Trust](https://hispecular.com/enterprise/security).

## Two-step sign-in

Turn it on for your own account in Settings → Security with an authenticator app, and keep the backup codes somewhere safe.

## Retention

In Settings → Data, choose how long closed conversations and uploaded files are kept: 30, 90, 180, 365 or 730 days, or until you remove them. The policy is applied every hour.

## Erasure requests

A person, a visitor, a conversation or the whole workspace can be erased on request. Requests wait 24 hours before they run, so a mistake can be cancelled from Settings → Data.


---

# Public API

> Read and write customers, deals, tickets, meetings and the live chat inbox over REST, with an OpenAPI document to build against.

- Canonical page: https://hispecular.com/docs/public-api
- Section: Developers
- Updated: 2026-10-22

The REST API reads and writes what you see in the workspace: customers, companies, leads, deals, tasks, notes, support tickets, meetings and the live chat inbox. It is included on Pro and Enterprise. Everything is JSON over HTTPS, and every endpoint is described in an OpenAPI document you can build against.

## What it covers

| Area | You can |
| --- | --- |
| [People](https://hispecular.com/docs/api-reference/people) | Create, read, update, search and archive customers. |
| [Companies](https://hispecular.com/docs/api-reference/companies) | The same for accounts. |
| [Leads](https://hispecular.com/docs/api-reference/leads) | Track and qualify people who have shown interest. |
| [Opportunities](https://hispecular.com/docs/api-reference/opportunities) | Deals: move them through a pipeline, win and lose them. |
| [Pipelines](https://hispecular.com/docs/api-reference/pipelines) | Read pipelines and their stages. |
| [Custom objects](https://hispecular.com/docs/api-reference/custom-objects) | Work the records of objects your workspace defines. |
| [Tasks](https://hispecular.com/docs/api-reference/tasks) | Create, change, complete and delete work for your team. |
| [Notes](https://hispecular.com/docs/api-reference/notes) | Add, edit, pin and delete notes on records. |
| [Tickets](https://hispecular.com/docs/api-reference/tickets) | Open tickets about a record and move them along. |
| [Meetings](https://hispecular.com/docs/api-reference/meetings) | Schedule, move and cancel meetings, with invitations. |
| [Inbox](https://hispecular.com/docs/api-reference/inbox) | Read live chats, take them over from the AI agent, reply, hand back. |
| [Team](https://hispecular.com/docs/api-reference/team) | List the people in the workspace, for owners and assignees. |
| [Webhooks](https://hispecular.com/docs/api-reference/webhooks) | Subscribe an address to events, instead of polling. |

## Quickstart

1. Open Settings → API and create a key with the scopes you need. Copy it straight away; it is shown once. See [Authentication and scopes](https://hispecular.com/docs/api-authentication).
2. Send it as a bearer token. This lists the most recently changed customers.

```bash
curl https://hispecular.com/api/v1/people \
  -H "Authorization: Bearer sk_live_YOUR_KEY"
```

1. Create a customer. The key needs the people write scope.

```bash
curl -X POST https://hispecular.com/api/v1/people \
  -H "Authorization: Bearer sk_live_YOUR_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: signup-form-2026-11-05-emily" \
  -d '{"display_name": "Emily Carter", "email": "emily@example.com"}'
```

The response is `{ "data": { "id": "pr_…", … } }`. Use that id to read, change or archive the customer, or to link a deal, task, note or ticket to them.

## Where to go next

- [Requests and responses](https://hispecular.com/docs/api-conventions): pages, filters, errors, retries and limits.
- [Webhooks](https://hispecular.com/docs/webhooks): know the moment something changes.
- [Recipes](https://hispecular.com/docs/api-recipes): common jobs, start to finish.
- [API reference](https://hispecular.com/docs/api-reference) and the [OpenAPI document](https://hispecular.com/api/v1/openapi.json) to generate a client, and a [Postman collection](https://hispecular.com/api/v1/postman.json) with every request ready to send.
- [MCP server](https://hispecular.com/docs/api-mcp): let an AI agent use the API.


---

# Authentication and scopes

> Create an API key, choose its scopes and role, act as a teammate, and keep keys safe.

- Canonical page: https://hispecular.com/docs/api-authentication
- Section: Developers
- Updated: 2026-10-22

Every request carries an API key. A key is a deliberate, narrow grant: it can do only what its scopes and its role allow, and every change it makes is recorded against it.

## Create a key

1. Open Settings → API. Only owners and admins manage keys.
2. Name the key for what it is used for, such as "Order sync".
3. Choose its scopes, its role, an optional expiry and its requests-per-minute limit.
4. Copy the key. It starts with `sk_live_`. We store only a hash, so a lost key cannot be recovered: revoke it and make another.

```bash
curl https://hispecular.com/api/v1/pipelines -H "Authorization: Bearer sk_live_YOUR_KEY"
```

A missing, wrong, revoked or expired key answers `401`. A workspace that is not on Pro or Enterprise answers `403` to every key.

## Scopes

A scope names a kind of access. A write scope includes the matching read. A call that needs a scope the key lacks answers `403` and names the scope it needed.

| Scope | Allows |
| --- | --- |
| `people:read` | Read customers |
| `people:write` | Create and change customers |
| `companies:read` | Read companies |
| `companies:write` | Create and change companies |
| `leads:read` | Read leads |
| `leads:write` | Create and change leads |
| `opportunities:read` | Read opportunities |
| `opportunities:write` | Create and change opportunities |
| `tasks:read` | Read tasks |
| `tasks:write` | Create and change tasks |
| `notes:read` | Read notes |
| `notes:write` | Create and change notes |
| `records:read` | Read custom object records |
| `records:write` | Create and change custom object records |
| `pipelines:read` | Read pipelines and stages |
| `evidence:read` | Read evidence |
| `reports:read` | Run reports |
| `webhooks:read` | Read webhook endpoints |
| `webhooks:write` | Manage webhook endpoints |
| `team:read` | Read teammates |
| `tickets:read` | Read support tickets |
| `tickets:write` | Open and change support tickets |
| `meetings:read` | Read meetings |
| `meetings:write` | Schedule, change and cancel meetings |
| `conversations:read` | Read live chat conversations |
| `conversations:write` | Reply in, assign and manage live chat conversations |

## Roles and acting as a person

Scopes are not the only gate. Every key also has a role, and a call must be something that role could do in the app. A key with the viewer role and a write scope can still not write, because a viewer cannot write in the app either. Keys can hold the admin, manager, member or viewer role, never owner.

A key can also **act as** one teammate. It then sees exactly what that teammate sees: their role, their teams and which records are visible to them. Fields hidden from that role are absent from responses, and records they cannot see answer `404`, as if they did not exist. Use this for a key that belongs to one person's automation.

Replies sent in the live chat inbox and meetings scheduled through the API are made as the person the key stands for: its creator, or the teammate it acts as.

## Keeping keys safe

- One key per job, with the fewest scopes it needs. Revoke a key the day it is not.
- Keep keys in environment variables or a secrets manager, never in a repository, a browser or a mobile app. A browser is not a safe place for a key.
- Set an expiry on keys given to contractors or trials.
- Every change a key makes is recorded with the key's name and prefix, so what a leaked key did can be traced.

> **If a key leaks** Revoke it in Settings → API straight away. Revoking takes effect on the next request. Then make a new key and review what the old one could reach.


---

# Requests and responses

> Pages, filters, syncing, error codes, retries with idempotency keys, and rate limits.

- Canonical page: https://hispecular.com/docs/api-conventions
- Section: Developers
- Updated: 2026-10-22

Every endpoint behaves the same way for pages, filters, errors, retries and limits. Learn these once.

## Responses

A success returns `{ "data": … }`: one object, or a list. Objects carry an `object` field naming what they are, a public `id` such as `pr_1a2b3c4d5e6f7a8b`, and timestamps in ISO 8601 with a zone. Every response has an `X-Request-Id` header; quote it if you contact support.

## Lists, pages and sync

Lists return `{ "data": [...], "pagination": { "nextCursor", "hasMore" } }`. Ask for up to 100 with `limit`, and pass `nextCursor` back as `cursor` for the next page. A cursor belongs to the filters it was made with; using it with others answers `422`.

To keep another system in step, ask for `sort=-updated_at` with `updated_since` set to the newest `updated_at` you last saw. Some lists cannot be paged and say so in the reference; they return the newest items and `hasMore`.

```bash
curl "https://hispecular.com/api/v1/people?sort=-updated_at&updated_since=2026-11-05T00:00:00Z&limit=100" \
  -H "Authorization: Bearer sk_live_YOUR_KEY"
```

## Filters, search and custom fields

- `q` searches a record's main text. Other query parameters filter by equality; each endpoint lists its own in the reference.
- Custom fields are written and read by their key beside the built-in ones. A value that does not fit the field's type is a `422` that names the field.
- Relations take public ids: a deal's `person` is a `pr_…` id. An id for something the key cannot see is a `422`, never a silent skip.

## Errors

Every failure has the same shape. Branch on `code`, never on the wording of `message`, which is written for people and may change.

```json
{
  "error": {
    "code": "validation_failed",
    "message": "The request body is invalid.",
    "requestId": "req_5b1f…",
    "details": [{ "field": "email", "message": "Enter a valid email address." }]
  }
}
```

| Status | Code | Meaning |
| --- | --- | --- |
| 401 | `unauthorized` | No key, or a key that is wrong, revoked or expired. |
| 403 | `forbidden` | The key lacks the scope, or its role cannot do this. |
| 404 | `not_found` | It does not exist, or this key cannot see it. |
| 409 | `conflict` | The state forbids it, such as a ticket move it cannot make. |
| 422 | `validation_failed` | Something in the request is wrong; details say what. |
| 422 | `idempotency_key_reused` | The key was used with a different request. |
| 429 | `rate_limited` | Too many requests. Wait for Retry-After. |
| 500 | `internal` | Our side. Retry; quote the request id if it persists. |

## Retries and idempotency

Send an `Idempotency-Key` (any unique string up to 255 characters) on creates a retry could repeat. Retrying with the same key and body replays the first answer and sets `Idempotent-Replayed: true`; the same key with a different body is refused. Keys are remembered for 24 hours. `GET`, `PATCH` and `DELETE` are safe to retry as they are.

## Rate limits

Each key has a requests-per-minute limit, set when it is made. Every response carries `RateLimit-Limit`, `RateLimit-Remaining` and `RateLimit-Reset`. Past the limit the API answers `429` with `Retry-After` in seconds; wait that long, then continue. Back off rather than retrying in a tight loop. Sending to people has its own, lower limits, noted where they apply: replies in the live chat inbox, and invitations sent by meetings.


---

# Webhooks

> Receive signed events when records change, subscribe from code, and verify each delivery.

- Canonical page: https://hispecular.com/docs/webhooks
- Section: Developers
- Updated: 2026-10-22

Webhooks send an HTTPS request to your endpoint when something changes in the workspace, so you do not have to poll. They are included on Pro and Enterprise.

## Add an endpoint

1. Open Settings → Webhooks and choose **Add an endpoint**.
2. Enter your HTTPS address and choose the events to send.
3. Copy the signing secret. You need it to verify deliveries.

A workspace can have up to 100 endpoints. An address that resolves to a private network is refused, and in production the address must be HTTPS.

## Subscribe with the API

A program can manage its own subscriptions, which is how an integration turns an instant trigger on and off. The key needs the `webhooks:write` scope and a role that can manage integrations.

```bash
curl -X POST https://hispecular.com/api/v1/webhooks \
  -H "Authorization: Bearer sk_live_YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{"url": "https://example.com/hooks/syncon", "events": ["record.created", "opportunity.won"]}'
```

The response includes `data.secret`, once. Store it with the subscription's `id`. To stop, `DELETE https://hispecular.com/api/v1/webhooks/{id}`. You can also list and change subscriptions, rotate a secret, and read a subscription's delivery history; see the [webhooks reference](https://hispecular.com/docs/api-reference/webhooks).

## Events

| Event | Sent when |
| --- | --- |
| `record.created` | A customer, company, lead, opportunity, quote or custom record was added. |
| `record.updated` | A field on a record changed. Carries the field key, not the value. |
| `record.deleted` | A record was permanently deleted. |
| `opportunity.stage_changed` | An opportunity moved to another open stage. |
| `opportunity.won` | An opportunity moved to a won stage. |
| `opportunity.lost` | An opportunity moved to a lost stage. |
| `task.completed` | A task was marked done. |
| `conversation.created` | A visitor or customer started a new conversation, on the website or a connected channel. |
| `conversation.message_received` | A visitor or customer sent a message. The text is not in the event; read the conversation. |
| `conversation.handoff_requested` | A visitor asked for a person, or left a follow-up for one. |
| `meeting.booked` | A visitor booked a meeting in a conversation. |
| `task.created` | A task was added. |
| `note.created` | A note was written on a customer, company, lead or deal. The text is not in the event; read the note. |

`GET https://hispecular.com/api/v1/webhooks/events` lists the events a key can subscribe to, including ones that are listed but not sent yet.

## What a delivery contains

A delivery is deliberately thin: identifiers and the kind of change, never field values. That keeps a webhook from becoming a way around field security. To get the record, read it with the API using a key that is allowed to see it.

```json
{
  "id": "evt_9c1e…",
  "type": "record.created",
  "createdAt": "2026-11-05T10:00:00.000Z",
  "organizationId": "org_…",
  "data": {
    "objectType": "person",
    "recordId": "6f1c…",
    "publicId": "pr_1a2b3c4d5e6f7a8b"
  }
}
```

`data.publicId` is the id the API uses, so `GET https://hispecular.com/api/v1/people/pr_1a2b3c4d5e6f7a8b` returns the customer. Updates add `fieldKey`; stage events add the stage facts. Conversation events carry `data.conversationId` (`GET https://hispecular.com/api/v1/conversations/conv_…`, with the latest message), a booking carries `meetingId`, and a new task carries `taskId`, each the id the API reads it by. What a visitor wrote is never in the event. A record that key cannot see answers `404`, which your handler should treat as nothing to do. Headers: `x-syncon-event`, `x-syncon-delivery` and `x-syncon-signature`.

## Verify a delivery

The `x-syncon-signature` header looks like `t=1729590000,v1=…`: a Unix timestamp and an HMAC-SHA256, in hex, of the timestamp, a full stop and the raw request body, keyed with your signing secret. Reject a delivery whose timestamp is more than five minutes old or whose signature does not match.

```javascript
import { createHmac, timingSafeEqual } from "node:crypto";

// body: the raw request body as a string, before any JSON parsing
export function isGenuine(body, header, secret) {
  const parts = Object.fromEntries(
    (header ?? "").split(",").map((part) => {
      const at = part.indexOf("=");
      return [part.slice(0, at).trim(), part.slice(at + 1).trim()];
    }),
  );
  const timestamp = Number(parts.t);
  if (!Number.isInteger(timestamp) || !/^[0-9a-f]{64}$/.test(parts.v1 ?? "")) return false;
  if (Math.abs(Date.now() / 1000 - timestamp) > 300) return false;

  const expected = createHmac("sha256", secret).update(`${timestamp}.${body}`).digest();
  return timingSafeEqual(expected, Buffer.from(parts.v1, "hex"));
}
```

```python
import hashlib, hmac, time

def is_genuine(body: bytes, header: str, secret: str) -> bool:
    parts = dict(p.split("=", 1) for p in (header or "").split(",") if "=" in p)
    try:
        timestamp = int(parts["t"])
        signature = parts["v1"]
    except (KeyError, ValueError):
        return False
    if abs(time.time() - timestamp) > 300:
        return False
    expected = hmac.new(
        secret.encode(), f"{timestamp}.".encode() + body, hashlib.sha256
    ).hexdigest()
    return hmac.compare_digest(expected, signature)
```

> **Verify against the raw body** Parsing the JSON and serialising it again changes the bytes, and the signature will not match. Read the body as text first.

## Delivery, retries and ordering

- Answer with any `2xx` within 10 seconds. Anything else, or no answer, is a failure.
- A failed delivery is retried with growing gaps, up to eight attempts. The same event keeps the same `id` on every attempt, so dedupe on it.
- Deliveries are not guaranteed to arrive in order. Use `createdAt`, or read the record for its current state.
- An endpoint whose deliveries keep failing every retry is switched off, and the workspace is told. Fix the address, then set its status back to `active`.
- Events typically reach your endpoint within a few seconds of the change.


---

# Recipes

> Common jobs, start to finish: a customer and a deal, reacting to a win, tickets, meetings, syncing.

- Canonical page: https://hispecular.com/docs/api-recipes
- Section: Developers
- Updated: 2026-10-22

Common jobs, start to finish. Each one is a handful of calls; the [reference](https://hispecular.com/docs/api-reference) has every field.

## Add a customer and open a deal

```bash
# 1. the customer (people:write)
curl -X POST https://hispecular.com/api/v1/people -H "Authorization: Bearer $KEY" \
  -H "Content-Type: application/json" -H "Idempotency-Key: order-1042-customer" \
  -d '{"display_name": "Emily Carter", "email": "emily@example.com"}'
# → { "data": { "id": "pr_…" } }

# 2. the pipelines, each with its stage ids (pipelines:read)
curl https://hispecular.com/api/v1/pipelines -H "Authorization: Bearer $KEY"

# 3. the deal (opportunities:write)
curl -X POST https://hispecular.com/api/v1/opportunities -H "Authorization: Bearer $KEY" \
  -H "Content-Type: application/json" -H "Idempotency-Key: order-1042-deal" \
  -d '{"name": "Annual plan", "person": "pr_…", "stage": "STAGE_ID", "amount": 1200}'
```

## React the moment a deal is won

1. Subscribe to `opportunity.won` ([Webhooks](https://hispecular.com/docs/webhooks)).
2. Verify the signature, then read the deal with `GET https://hispecular.com/api/v1/opportunities/{data.publicId}`.
3. Do your work, and answer `200` quickly. Do slow work after you have answered.

## Open a ticket and resolve it

```bash
curl -X POST https://hispecular.com/api/v1/tickets -H "Authorization: Bearer $KEY" \
  -H "Content-Type: application/json" \
  -d '{"subject": "Invoice shows the wrong ABN", "person": "pr_…", "priority": "high"}'
# → { "data": { "id": "1042", "number": 1042, "status": "new" } }

curl -X PATCH https://hispecular.com/api/v1/tickets/1042 -H "Authorization: Bearer $KEY" \
  -H "Content-Type: application/json" -d '{"status": "resolved"}'
```

A ticket can only move along its lifecycle. A move it cannot make answers `409`. Tickets are closed, never deleted.

## Schedule a meeting

```bash
curl -X POST https://hispecular.com/api/v1/meetings -H "Authorization: Bearer $KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "title": "Quarterly review",
    "starts_at": "2026-11-12T10:00:00+11:00",
    "duration_minutes": 45,
    "time_zone": "Australia/Sydney",
    "person": "pr_…",
    "attendees": [{ "email": "emily@example.com" }],
    "notify": false
  }'
```

With `notify` left on, the meeting goes on the owner's connected calendar and the attendees are invited. Set it to `false` to save it without telling anyone. The response's `notice` says who was told.

## Answer a live chat

See [The live chat inbox](https://hispecular.com/docs/api-live-chat): join the conversation, reply, and hand it back.

## Keep another system in step

1. Do a first full read of the list, paging with `cursor`, and remember the newest `updated_at`.
2. Subscribe to `record.created`, `record.updated` and `record.deleted` and apply each change as it arrives.
3. On a schedule, and after any outage, read `?sort=-updated_at&updated_since=…` to catch anything a delivery missed.


---

# The live chat inbox

> Read conversations, take them over from the AI agent, reply, hand back, and organise them through the API.

- Canonical page: https://hispecular.com/docs/api-live-chat
- Section: Developers
- Updated: 2026-10-22

Everything a teammate does in the live chat inbox can be done through the API: read the queue and a conversation, take it over from the AI agent, reply to the visitor, hand it back, and organise it. It uses the same rules as the inbox in the app.

> **Replies reach real people** A reply is read by a visitor at once and cannot be unsent. Replies are made as the person the key stands for (its creator, or the teammate it acts as), so the visitor sees that person's name. Keys need the `conversations:write` scope and a role that can write in the inbox, and are limited to 30 replies a minute.

## Read the queue and a thread

```bash
# the queue (conversations:read)
curl "https://hispecular.com/api/v1/conversations?view=needs_person&limit=25" -H "Authorization: Bearer $KEY"

# one conversation: who is answering, and whether you may reply
curl https://hispecular.com/api/v1/conversations/conv_… -H "Authorization: Bearer $KEY"

# its thread, oldest first; pass the last cursor as "after" to follow it live
curl "https://hispecular.com/api/v1/conversations/conv_…/messages?after=0" -H "Authorization: Bearer $KEY"
```

`responder` is `ai` while the agent is answering and `human` once a teammate has taken over. Reading never marks a conversation read. Messages carry a `role` of `visitor`, `agent`, `human` or `system` (a join, a hand-back, a booking).

## Take over and reply

1. `POST /conversations/{id}/join`. The visitor sees a teammate arrive. If another teammate already has it, you get `409`.
2. `POST /conversations/{id}/reply` with `{ "text": "…" }`. Replying before joining is a `409`.

```bash
curl -X POST https://hispecular.com/api/v1/conversations/conv_…/reply -H "Authorization: Bearer $KEY" \
  -H "Content-Type: application/json" \
  -d '{"text": "Yes, we ship to Perth.", "client_message_id": "case-88-reply-1"}'
```

Send your own `client_message_id` so a retry after a dropped connection posts the message once, not twice.

## Hand back to the AI agent

`POST /conversations/{id}/return-to-ai` returns the conversation to the agent. Only the teammate who holds it can.

## Owner, priority, tags and notes

- `PATCH /conversations/{id}` sets `assignee` (a teammate from `GET /team/members`, or `null`), `priority`, and `add_tags` / `remove_tags`.
- `POST /conversations/{id}/notes` adds a note only your team sees; the visitor never does. `GET` lists them.
- There is no separate close. A conversation is resolved by resolving its ticket, with `PATCH /tickets/{number}`.

See the [inbox reference](https://hispecular.com/docs/api-reference/inbox) for every field.


---

# MCP server for AI agents

> Let Claude, ChatGPT, Cursor or your own agent work the workspace, limited to what its key allows.

- Canonical page: https://hispecular.com/docs/api-mcp
- Section: Developers
- Updated: 2026-10-22

Specular runs an MCP (Model Context Protocol) server, so an AI agent such as Claude, ChatGPT, Cursor or an agent in your own workflow can work the workspace through the API. Every endpoint is a tool, and an agent can do only what its API key allows.

## Connect

The endpoint is `POST https://hispecular.com/api/v1/mcp`, with your API key as a bearer token. It speaks the Streamable HTTP form of MCP: one JSON message in, one JSON answer out, no session. Clients that take a URL and headers use it directly. A client that only speaks stdio can go through a bridge:

```bash
npx mcp-remote https://hispecular.com/api/v1/mcp --header "Authorization: Bearer sk_live_YOUR_KEY"
```

## Tools

A tool is named after its operation, such as `list_tickets`, `create_person`, `join_conversation` or `reply_to_conversation`. The list an agent sees is narrowed to the scopes its key holds, so it is not offered what it would be refused. Each tool says whether it only reads, and which deletes cannot be undone. The names and fields match the [reference](https://hispecular.com/docs/api-reference).

## Keeping an agent safe

- Give the agent its own key with the fewest scopes that do the job. Read-only scopes cannot change anything.
- Text in records, notes, tickets and conversations was written by customers and other third parties. The server tells the agent to treat it as data and never as instructions; still, do not give an agent that reads untrusted text a key that can delete or send.
- A tool call is an ordinary API request with the same key, so scopes, roles, record visibility, rate limits and the record of changes all apply, and each change names the key.
- Sending a reply or scheduling a meeting reaches real people. Ask the agent to show you the wording and wait for your go-ahead.


---

# Connect an automation tool

> Use Zapier, Make, n8n, Pabbly or similar with HTTP and webhook steps, or build Specular into your own product.

- Canonical page: https://hispecular.com/docs/api-automation-tools
- Section: Developers
- Updated: 2026-10-22

You can connect Specular to an automation tool today with the tool's own HTTP-request and webhook steps, using the same API and webhooks as any program. Zapier, Make, n8n, Pabbly Connect, Activepieces and similar tools all have them.

## Sign in

Add a header to each HTTP step: `Authorization: Bearer sk_live_YOUR_KEY`, with the base address `https://hispecular.com/api/v1`. Make one key for the tool, with only the scopes its flows need.

## Start a flow when something happens

Use the tool's webhook trigger to get an address, then subscribe that address with `POST /webhooks` and the events you want (see [Webhooks](https://hispecular.com/docs/webhooks)). A delivery carries only ids, so add a step that reads the record: for a new customer, `GET /people/{data.publicId}`. When the flow is switched off, delete the subscription.

If a tool cannot register webhooks, poll instead: read a list newest first, for example `GET /people?sort=-created_at&limit=50`, and skip the ids it has seen.

## Do something in Specular

Each action is one HTTP step: create a customer, add a note, open a ticket, schedule a meeting, reply in a chat. The [reference](https://hispecular.com/docs/api-reference) gives the path, method and body for each. Send an `Idempotency-Key` on creates so a retried step does not make a second record, and show the API's `error.message` when a step fails.

## Build your own integration

If you make an automation product, you can build Specular into it as an app: customers paste their key as the sign-in, your triggers subscribe webhooks, and your actions call the endpoints. Generate a client from the [OpenAPI document](https://hispecular.com/api/v1/openapi.json) rather than writing requests by hand.