Trazio by use case

Pick the thing you want to watch. Each guide follows the same path: send data, check it landed, see which rules fire, then ask about it from an AI assistant, the SDK or curl.

I want to watch…	Data it takes in	Rules that fire out of the box	Guide
Robots and autonomous machines (state, faults, battery, ROS 2 / MCAP recordings)	`robot.`, `vehicle.`, ROS 2, MCAP files	`physical-ai`	Robotics telemetry
Cameras and robot video (who is near what, in a live or recorded scene)	live camera frames, recorded video	`live-physical-safety`, `live-robot-human-interaction`	Robot and camera video
Security (endpoint, SIEM, authentication, syslog)	Splunk HEC, Loki, Elasticsearch Bulk, NDJSON, syslog	`security-endpoint`	Cybersecurity
Cloud and Kubernetes (workloads, deployments, control plane)	`kubernetes.*`, Argo CD, OpenTelemetry	`cloud-kubernetes`	Cloud and Kubernetes
Application logs, metrics, traces and AI agents	OpenTelemetry, Fluent Bit, Vector, Alloy	`observability`	Application observability

New here? Start with Getting started, which covers what you need before the first event and how to connect an assistant. To create your own pipelines, follow The guided setup, or see Create pipelines from the console, code or an assistant.

Ways to send and to ask

	curl	Python	Node.js	Collector
Send	any `POST /v1/ingest`	`pip install trazio-telemetry`	`npm install trazio-telemetry`	OpenTelemetry Collector, Grafana Alloy, Fluent Bit, Vector
Ask	`GET /v1/query`	`client.semantic_query(...)`	`client.semanticQuery(...)`	n/a

Assistants (Claude, ChatGPT and other MCP clients) ask the same questions in plain language at https://ca.mcp.trazio.cerebrixos.com/mcp.

What is checked

The commands in these guides come from the operator-tested runbooks in Tenant operations, section 5. The assistant answers marked Sample were produced against the shared trazio-demo pipeline.