Open-Source AI Hedge Fund Replicates Investment Philosophies for Retail Investors
EP
Elliot Pemberton
Warren Buffett · Apr 14, 2026
Source: DojiDoji Data Terminal
Retail investors can now access synthesized buy, sell, or hold signals based on the replicated investment logic of 12 world-class investment experts. The AI Hedge Fund project, open-sourced by independent developer Virat Singh on GitHub, encodes the distinctive stock-picking logic and risk preferences of investors such as Warren Buffett, Charlie Munger, and Ben Graham into AI agents.
To generate these signals, the system employs a three-tier architecture featuring a React 18 frontend and a Python FastAPI backend. The frontend integrates a React Flow visual workflow editor, allowing users to drag and drop agent nodes to design their own investment committee. The backend orchestrates multi-agent workflows using LangGraph, while the data layer connects to external APIs for real-time market data and financial statements.
Deployment is available via command-line or web application, and the system is compatible with 13 large language models, including OpenAI, Anthropic, Groq, and DeepSeek. For users seeking to run inference without an internet connection, the system supports local models via the —ollama parameter. Users can validate strategies before deploying real capital by running the backtesting module on historical data.
Warren Buffett
The Ledger Morning
The essential intelligence to start your trading day. Delivered 6:00 AM EST.
Join 50,000+ professionals who start their day with The Digital Ledger.