Your Voice Agents home (~52s)

Tour the agents list: folders, search, filters, views, and the Create Agent button.

Organize agents with folders (~45s)

Create, rename, and delete folders, move agents between them, and archive retired agents.

Create an agent from a template (~45s)

Pick a production-ready template, preview its flow, and spin up your agent in one click.

Start blank or import JSON (~41s)

Start from an empty agent with an optional seed prompt, or import a previously exported flow.

Describe the job, let AI build it (~44s)

Describe the call you want and let the AI Builder draft a complete agent flow for you.

The single-prompt editor (~57s)

Get to know the three-column editor: prompt, settings rail, and live tester.

Write the prompt & welcome message (~62s)

Shape your agentтАЩs persona and rules in the prompt, and control exactly how calls open.

Voice, models & speech settings (~65s)

Choose voices and models, and tune speech settings, turn detection, and transcription.

Knowledge, extraction & webhooks (~57s)

Attach a knowledge base, extract structured data after calls, and send events to webhooks.

Tools, MCPs & call controls (~56s)

Wire functions and MCP servers into your agent, and set call limits, security, and fallbacks.

Test your agent in the browser (~57s)

Run live audio and chat tests against your agent without leaving the editor.

The conversational-flow canvas (~64s)

Build deterministic conversation flows visually with nodes, edges, and a live preview.

Configure nodes & transitions (~74s)

Edit node settings, connect transitions, and control how the conversation moves.

Review runs and call details (~60s)

Inspect every run your agent handles, with analytics up top and full call detail.

The agent settings page (~46s)

Manage long-tail options like voicemail detection, ambient noise, and embedding from one page.

Embed your agent on a website (~53s)

Turn on the web widget, lock it to your domains, and drop it into your site.