This repository provides a Financial RAG system for ingesting financial documents and generating grounded answers. It features PDF and URL ingestion, hybrid retrieval with BM25 and vector search, and a reranking pipeline. The system uses FastAPI, Next.js, PostgreSQL, and Groq LLM API.
From the README
FinRAG — Financial Research Assistant A production-style Financial RAG (Retrieval-Augmented Generation) system that allows users to ingest financial documents, retrieve relevant context, and generate grounded answers using LLMs. Built using: • FastAPI • Next.js • PostgreSQL • BM25 + Vector Search • Groq LLM API • Hybrid Retrieval + Reranking --- Features • Upload and ingest financial PDFs • Ingest financial URLs/articles • Hybrid retrieval: • BM25 lexical search • Dense vector embeddings • Reranking pipeline • Grounded financial Q&A • Groq-powered fast inference • Evaluation pipeline • Next.js modern frontend • Swagger API docs • PostgreSQL persistence --- Tech Stack Frontend •...
Skills
- Python