#ai-observability

企业级 AI 资源纳管平台，提供统一 AI网关、Token调度能力，纳管 OpenAI、Azure、Claude、DeepSeek 等主流模型，并支持 MCP 工具与 Skill 的集中注册分发。具备内外双轨定价、成本归因、统一身份认证、安全审计与效能报表，帮助企业精准控制成本、量化 ROI，高效治理 AI 资产。

Deterministic health monitoring for LLM conversations. Detects context pressure, drift, repetition and tool anomalies without adding anything to model context.

A practical 12-week Agentic AI learning roadmap covering LLM foundations, tool calling, RAG, memory, MCP, multi-agent systems, evaluation, security, and production deployment.

Open-source AI Operations platform for multi-agent systems featuring OpenTelemetry instrumentation, SigNoz observability, MCP-powered diagnostics, chaos engineering, workflow replay, and AI cost intelligence.

Local-first OpenTelemetry and Grafana diagnostic kit for Codex, with Loki logs, Tempo traces, Prometheus spanmetrics, privacy-safe derived evidence, and focused Grafana dashboards.

Production-inspired AI observability lab: a traced and evaluated fictional banking RAG assistant with OpenTelemetry, Phoenix, prompt experiments, guardrails, and optional Dynatrace fan-out.