virattt/ai-hedge-fund
An AI Hedge Fund Team
What it solves
It provides a proof-of-concept framework for exploring how AI can be used to make trading decisions. It allows users to build, backtest, and paper-trade simulated hedge funds without risking real capital.
How it works
The system uses LLM-powered "alpha models" to analyze financial data (prices, fundamentals, and earnings) sourced from the Financial Datasets API. Users define a "mandate"—which specifies strategies, risk parameters, and rebalance cadence—and then apply that mandate to specific stock tickers to simulate trading cycles.
Who it’s for
It is designed for researchers and learners interested in the intersection of AI and quantitative finance who want to experiment with AI-driven investment strategies in a simulated environment.
Highlights
- Multi-LLM Support: Compatible with providers including Anthropic, OpenAI, DeepSeek, Google, xAI, and Kimi.
- Backtesting: Ability to run mandates over historical data to visualize equity curves against benchmarks.
- Mandate-based Configuration: Separates the fund's strategy and risk settings from the specific assets being traded.
- Interactive Terminal App: Includes a CLI for building and managing funds.
Related
- Project
- Project
- Project
- Project
- Project