This project generates real-time trading insights for stock markets by analyzing core market data, fundamental data, corporate events and external context using a quantized large language model (LLM) and Retrieval-Augmented Generation (RAG) pipeline.
- Real-time data ingestion from Yahoo Finance
- Vector database (ChromaDB + FAISS) for efficient retrieval
- Quantized LLM (LLaMA or similar) via
llama-cpp-pythonfor local inference - RAG pipeline for contextual insight generation
- Output displays the insights behind market movements