ufund-me/qbot

Qbot is a Chinese-developed, free and open-source quantitative investment platform with a heavy emphasis on AI. In plain terms, it is a toolkit for building, testing, and running automated trading bots, programs that decide when to buy and sell financial instruments, and it focuses on using machine learning rather than only hand-written rules. The README's diagram summarises it as intelligent trading strategies plus a backtesting system plus automated trading, with visualisation tools on top. The way it works is split into modular layers: a data layer pulls in market data, a strategy layer is where the user writes or trains the trading logic, including AI strategies built with machine learning, reinforcement learning, or deep learning combined with multi-factor models, and a trading engine connects to the market to execute orders. The project stitches together several existing Python libraries: qlib for AI strategy research, backtrader and easyquant for backtests, vnpy and pytrader for live order routing, and quantstats for performance dashboards. It supports stocks, funds, futures, and cryptocurrencies, and ships with a GUI client so non-developers can drive it. Someone would use this if they want a fully local, self-hosted way to evaluate stock or fund ideas, run historical backtests, paper-trade in a simulated environment, and eventually flip a switch to automated live trading on a single platform. The author notes that a little Python knowledge and a little trading experience help. The code is Python in Jupyter notebooks and has been tested on Python 3.8 and 3.9. Documentation lives at ufund-me.github.io/Qbot. The full README is longer than what was provided.