NO. 001

AIOps-example

AI engineering architecture pattern showcase (RAG · LangGraph · MCP · sandboxed execution)

AI
Type Architecture showcase
Status Actively evolving
Tech LangGraph · LangChain · FastMCP · Qdrant
Observability Langfuse · Prometheus · Grafana

AIOps-example is an enterprise-grade showcase of AI engineering architecture patterns, set in a simulated IT operations/SRE domain (runbooks, historical tickets, alerts, change records). Each top-level directory is a fully runnable example of one pattern — from basic RAG, to advanced hybrid retrieval, LangGraph multi-agent systems (Adaptive RAG, human-in-the-loop approval, Supervisor routing), MCP tool-protocol integration, agent orchestration comparisons (ReAct / Plan-and-Execute / Planner-Generator-Evaluator), an enterprise API gateway (RBAC, rate limiting, audit logging, multi-tenancy), a side-by-side multi-agent framework comparison (LangGraph vs Smolagents vs CrewAI vs AutoGen), to a sandboxed code-execution agent. Every LLM/agent call is traced with Langfuse, and service metrics are monitored via Prometheus + Grafana. All data is deterministically generated mock data — nothing connects to a real production system.

PythonLangGraphLangChainFastMCPFastAPIQdrantLangfusePrometheusGrafanaDockeruv