Internal Hackathon for Inter-IIT Tech 15.0
FinAgent
A fault-tolerant, RAG-augmented multi-agent system powered by openai/gpt-oss-120b on Groq. It ingests financial documents and live market signals, processing them through a LangGraph state machine to synthesize structured, observable trading decisions.
Tech Stack
LLM Engine: openai/gpt-oss-120b (via Groq)
Embeddings: all-MiniLM-L6-v2 (via HuggingFace)
Orchestration: LangGraph & LangChain
Infrastructure: FAISS, yfinance, Pandas, Streamlit
Quickstart
Clone the repository and navigate to the directory:
git clone https://github.com/YourUsername/FinAgent.git cd FinAgent
Install dependencies:
pip install -r requirements.txt
Run the application:
streamlit run app.py
Usage
Provide your Groq API Key in the sidebar. Upload contextual financial documents, then input live tickers or upload a CSV batch to trigger the analysis pipeline.