How to integrate intents to agents and analytics

Guides for agents, analytics export, and instrumentation — pick a topic below.

Agent layer

Feed behavioral intent into your AI agent — SDK callback, inference pull API, agent prompts, and webhooks.

  • Feed intent into your agentEnd-to-end tutorial — __ILM_ON_INTENT__ cache, chat-open prompt fetch, backend proxy, webhooks, and LLM wiring.
  • Agent layer API referenceInference and Config API endpoints — POST /v1/analyze, GET /v1/intents/latest, agent-prompts, curl and Python examples.

Analytics

Export Overview dashboard metrics for BI, warehouses, and scheduled sync jobs.

  • Analytics exportDownload Overview dashboard data as JSON — from the Export button or the Config API with curl, Python, and cron examples.

Instrumentation

URL patterns, in-app views, opaque logged-in identity, and trusted server-side events.

  • Server-side eventsStripe, Auth0, and Supabase webhook templates for login success, purchases, and other tokens that must not come from the browser.
  • Logged-in user identityOpaque user_id for CRM stitch and pull APIs — setUserIdentity, email-only HMAC recipe, and what works with visitor/session only.
  • Setup troubleshooting FAQEmpty Insights, consent accepted but no tokens, dead inference host, CORS, SPA views, and rediscovery after launch.

Pricing

Compare intentLM to in-house and third-party alternatives.

Compare costModel intentLM pricing against in-house LLM intent pipelines and third-party analytics.