feat: Update Dockerfile to force use of locally cached model and skip HuggingFace network calls at runtime

docs: Update README with new architecture diagrams, refined output examples, and revised latency figures.

feat: implement a metrics collection system for RLM queries and integrate it into the RLM pipeline.

feat: Add Qdrant vector store service for collection management, chunk upsertion, and similarity search.

feat: Add chat API endpoint with recursive sub-LLM pipeline and vector search fallback.

feat: Update requirements.txt to remove specific version numbers and lets pip decide

feat: Update text chunking service with configurable parameters and its dependencies.

build: update requirements.txt to specify project dependencies.

feat: Update `requirements.txt` to specify langchain-community dependencies version.

Init