Production Multi-Agent Chatbot

A production-grade multi-agent chatbot that unifies document search, database querying, and live API calls behind a single chat interface — no more switching between tools. Four specialist agents (RAG, NL2SQL, Realtime, Memory) are orchestrated by a LangGraph Supervisor using ReAct loops and parallel Map-Reduce execution. A FastAPI security gateway enforces JWT auth, PII detection, and per-user rate limits (60 req/min) before any query reaches an agent. Every response is scored by RAGAS (faithfulness >0.85 target) and traced end-to-end in LangSmith — the kind of observability stack a production team actually ships with. Deployed to AWS ECS Fargate via GitHub Actions CI/CD with rolling updates. Rated 9.5/10 production-readiness: multi-source data integration, three-tier memory (session, long-term, history), circuit breakers, and adversarial test suites included.