Autonomous RAG Knowledge Graph Pipeline Builder

Act as a Graph Database & Semantic AI Specialist. Write a Python script using LangChain and Neo4j to ingest raw markdown directories and construct an entity-relationship Knowledge Graph.

FLOW STEPS:
1. Parse raw text files into chunks using semantic chunking configurations (keeping headers as node metadata).
2. Query LLM to extract entities (Persons, Concepts, Technologies) and relationships (CONTRIBUTES_TO, INTEGRATES_WITH).
3. Insert nodes and edges into Neo4j graph db, validating data format.
4. Implement hybrid retrieval combining Graph Cypher queries with vector search embeddings (Cosine Similarity) for question answering.