CorpMind — Enterprise Knowledge Intelligence

Completed
2026

Multi-agent RAG platform that unifies enterprise knowledge across Gmail, Drive, Notion, Slack, and PDFs into a single intelligent interface. A LangGraph supervisor routes queries to 11 specialized agents, retrieves context via hybrid vector (Qdrant) + graph (Neo4j) search, detects contradictions with DeBERTa NLI, and returns cited confidence-scored answers. Implements 6 RAG patterns: Naive, Retrieve-and-Rerank (CrossEncoder top-20→5), Graph RAG, Hybrid RAG, Agentic RAG, and Multi-Agent RAG.

CorpMind — Enterprise Knowledge Intelligence - Image 1
About This Project

Multi-agent RAG platform that unifies enterprise knowledge across Gmail, Drive, Notion, Slack, and PDFs into a single intelligent interface. A LangGraph supervisor routes queries to 11 specialized agents, retrieves context via hybrid vector (Qdrant) + graph (Neo4j) search, detects contradictions with DeBERTa NLI, and returns cited confidence-scored answers. Implements 6 RAG patterns: Naive, Retrieve-and-Rerank (CrossEncoder top-20→5), Graph RAG, Hybrid RAG, Agentic RAG, and Multi-Agent RAG.

Technologies Used
LangGraphFastAPINext.jsQdrantNeo4jDeBERTaGroq LLaMAPython
Resume
CorpMind — Enterprise Knowledge Intelligence — Project