microsoft/qlib ? reverse-engineered prompt

Build me a Python-based platform for quantitative trading that uses AI to find and execute investment strategies. The main goal is to create a complete system that covers the whole process, from research to production.

The platform should be able to handle and process large amounts of financial data, like stock prices. It needs to allow users to build, train, and backtest different kinds of machine learning models to see how well their strategies would have performed.

A key feature should be an advanced AI agent that can automatically research and discover new trading "factors" or market signals. This agent should be able to optimize the trading models on its own over time. Basically, I want an end-to-end tool that empowers people to use AI for financial research and algorithmic trading, from exploring ideas to implementing them.