Docs
Introduction
Introduction
AITracer documentation for traceability, governance, verification, and AI cost visibility.
AITracer is the visibility layer for production AI calls. It turns black-box LLM and agent executions into secure, auditable, and cost-aware records.
The product identity comes from the original trace explorer experience:
- Prove every response with canonical records and SHA-256 integrity hashes.
- Track every penny with token usage, model distribution, and action-level cost.
- Secure every prompt with high-risk heuristics and governance policy results.
- Monitor performance with P95 latency and anomaly signals.
AITracer is designed for enterprise governance, observability, and verification across modern AI systems.
Documentation scope: how to integrate your apps and models, call the HTTP APIs, use SDKs, and apply common self-hosted or cloud deployment patterns. It does not describe vendor-internal hosting topology, operational secrets, or proprietary infrastructure diagrams.