Skip to main content

Menace Voice documentation

Build a voice agent, connect your systems, and operate every conversation with confidence.
2 min read

Build conversational agents that answer phone calls, qualify leads, support customers, and talk to visitors on your website. Menace Voice connects a visual conversation workflow to speech models, tools, knowledge, and your telephony provider.

Start with one successful browser conversation. Then publish the reviewed agent, connect a delivery channel, and use the run record to verify the result.

The path to production#

StageWhat you doWhat to verify
ConfigureChoose hosted or self-hosted, then connect your model providersModel validation succeeds in the target workspace
BuildDefine the conversation and its toolsThe draft has a complete path to an end node
TestRun text tests and a Web CallCorrect answers, tool arguments, audio, and failure behavior
PublishRelease the reviewed draftThe expected definition is the published version
ConnectAdd a phone route, outbound trigger, or website embedA completed call reaches the intended agent
OperateReview runs, costs, reports, and delivery to your systemsThe transcript and downstream business result agree

Find the right reference#

  • Start explains the platform and the first working conversation.
  • Build covers workflow design, models, voice, nodes, and tools.
  • Connect covers telephony providers, web deployment, integrations, and campaigns.
  • Operate covers release checks, run review, billing, troubleshooting, and self-hosting.
  • Developers covers data contracts, SDKs, and AI coding assistants.
  • API Reference documents the endpoints in the checked-in OpenAPI specification.
Note

The custom in-app guides remain available. Their builder and coding-agent handbooks are also included here from the same source.