Architecture
How OpenTelemetry and runtime context become production intelligence.
Docs v1.0Updated Supported
Architecture
AutoObserve is organized as a production-intelligence system, not a dashboard catalog:
Production
Telemetry · Changes · Topology · Context
│
▼
Evidence
│
▼
Production Intelligence
Detect → Investigate → Understand
│
▼
Decision → Response ↺
│
▼
Multi-DSL investigation engine
(CIR · Knowledge · Policy)
Signals become evidence. Evidence becomes understanding. Understanding becomes a decision.
The Platform page walks this model end to end — evidence provenance, the intelligence canvas, explainable decisions, response maturity, and how AutoObserve coexists with the stack you already run.
AutoObserve is OpenTelemetry-native and Kubernetes-aware. It is designed to be inspectable: every claim in an investigation should lead back to evidence and raw telemetry.