What is Callab AI
Callab AI runs AI voice and text agents that answer and place calls on your existing phone system, and chat on your website and messaging apps.
Callab AI is a platform for building AI agents that talk to your customers. A voice agent answers or places phone calls. A text agent replies on your website, in Slack, Telegram, Microsoft Teams, WhatsApp or Discord. Both run from one dashboard, and both are built the same way: a prompt, a voice or a model, optional tools, and a channel to reach people on.
You do not need to change your phone system. Agents connect to the numbers you already have through a SIP trunkSIP trunkA carrier connection that brings your own phone numbers to Callab AI.See the glossary, your Twilio account or a Go Auto Dial extension.
What an agent can do
- Answer inbound calls on a phone number and resolve the reason for the call: book an appointment, look up an order, answer a question from your documents.
- Place outbound calls to a list of contacts as a campaign, with retries, calling windows and a per-contact outcome.
- Hand off to a person by transferring the call, or to another agent.
- Act during the call through tools: send an email, call a webhook, schedule a meeting on Cal.com, push data to Zapier, or query an MCPMCPModel Context Protocol. A tool type that connects an agent to an MCP server's tools and data.See the glossary server.
- Capture structured outcomes after every conversation: intent, sentiment, the fields you define, and a summary and quality score for each call.
How the platform is organised
The dashboard sidebar is grouped by area, and the groups follow the life of an agent. Dashboard sits above them all.
| Group | Section | What lives there |
|---|---|---|
| AI & Knowledge | AI Agents | Every agent, and the workspace where you build, test and release one. |
| Knowledge Bases | Documents your agents can search during a conversation. | |
| Deployments | Launch AI Voice | Campaigns: inbound and outbound calling with phone numbers and contacts. |
| Launch AI Text | Deployments of text agents to messaging platforms. | |
| Observability | Call Logs, Conversations | Every call and every text thread, with transcript, recording and outcomes. |
| Advisor | Configuration issues detected across your agents, knowledge bases, campaigns and numbers. | |
| Communication | Phone Numbers, Contacts | The numbers agents call from and the people they call. |
| Connections | Integrations | CRM connections, automations and webhooks. |
| Configuration | Settings | Profile, members, workspaces, billing, storage and security. |
Opening Settings or an agent replaces the sidebar with that section’s own menu; Back to app at the top returns to this one.
What’s new, the megaphone in the header, lists recent changes to the product. It is scoped to where you are: on a page it shows the changes for that area, and on Dashboard it shows everything.
The documentation follows the same order. Build covers the agent workspace, Test and improve the checks before release, Deploy the channels, Operate what happens once calls flow, and Account and platform the rest.
Where to start
Quickstart: your first voice agent
Create an agent, give it a prompt and talk to it in the browser. About ten minutes.
Quickstart: your first text agent
The same for chat, ending with a widget you can put on a page.
Core concepts
Agents, versions, campaigns, outcomes and the other words the dashboard uses.
Anatomy of an agent
Every section of the agent workspace and what it controls.