These four playbooks mirror how teams typically adopt Inquir Compute: AI agent backends, webhook processors, cron jobs, and REST API endpoints. Each playbook pairs serverless functions with the HTTP gateway, environment secrets for provider credentials, and observability for when something misbehaves. Follow the links to the public use-case hubs for verification patterns, retries, and templates.
AI agent backends
Run long model calls, tool loops, and structured outputs behind a stable HTTP route. Enable streaming when you need partial tokens, and keep provider keys in environment secrets.
Webhook processors
Verify signatures on the raw body, return a 2xx response quickly, and offload heavy work to async jobs or pipelines when the provider enforces a short timeout.
Cron jobs
Run scheduled functions for syncs, reports, and cleanup. They use the same runtime and secrets as online handlers; write idempotent updates and use observability to audit each run.
REST API endpoints
Serve route groups and authenticated JSON APIs on the gateway. Parse event.body as a string on HTTP invokes, and share types between handlers where it helps.