Overview
Use Platform APIs to create and configure the AI agents behind your customer experiences. Manage Agents, Playbooks, Workflows, Knowledge and tools, then publish changes and inspect test and production results.
To connect a customer-facing interface to an agent, use Chat agents or Voice agents. These sections cover conversation sessions, customer input and agent responses.
Make your first Platform API request
Choose a personal or service key, then call the resource you want to work with. This example lists the workspaces available to your credential:
The Workspaces reference defines the request, response and access requirements. Each resource reference owns its operations, required fields and documented errors.
Manage agent resources
Create and configure Agents. Manage their Playbooks, Workflows, tools and variables through the related resource references.
Organize projects and workspaces, and manage project resources.
Manage Environments and publish agent changes.
Send REST test turns and run test scripts before releasing changes.
Connect Knowledge and tools
- Knowledge base queries: search project documents directly.
- Documents: add and manage the content an agent uses.
- Tools: manage the tools available to an agent.
Review results and usage
- Transcripts: retrieve conversations and their details.
- Evaluations: manage criteria and score conversations.
- Analytics: query usage and cost across conversations, models and resources.
Automate platform workflows
- Webhooks: receive session lifecycle and organization events at your own receiver.
- CLI: build and manage agents from a terminal or script.
- MCP Servers: connect an AI coding tool to Voiceflow.