Agent Configuration
| Guide | Description |
|---|---|
| Writing Effective Instructions | Define your agent’s role, goal, context, constraints, and output format. Includes a prompt framework, sample prompt, and common issues to avoid. |
Data and Context
| Guide | Description |
|---|---|
| Use Metadata Fields Passed from the API | Pass custom metadata via the Execute API and reference it in agent prompts using session memory. Includes schema details and prompt syntax examples. |
| Pass Files Uploaded by Users to External Services or LLMs | Access file metadata stored in session memory and forward file URLs or content to external systems or LLMs for further processing. |
| Structured Response Rendering via Artifacts | Render structured response using channel-specific components instead of textual response. |
| Configure Transcription | Configure transcription for logging, analytics of voice interactions. |
AI for Service Integration
| Guide | Description |
|---|---|
| Pass Contextual Data from AI for Service to Agent Platform | Share variables such as channel information and user preferences from the AI for Service Automation Node to the Platform’s session memory. |
| Share Contextual Metadata from Agent Platform to AI for Service | Configure event-driven handoffs that pass conversation history, user data, and system metadata from the Platform back to AI for Service. |
| Initiate a Live Agent Handoff from Agentic Apps | Configure the Agent Handoff Event and set up the AI for Service Automation Node to route conversations from an AI agent to a live agent. |
| Voice Interactions | Configure Real-Time Voice or TTS Streaming for voice channels integrated through the AI for Service Automation Node. |