Didact AI: The anatomy of an ML-powered stock picking engine
I started Didact AI in 2019, out of a firm belief that stock picking engines built on top of machine learning models could consistently beat the market, gaining and retaining profits.
I worked on this idea full-time for 2 years, from 2019 to 2021, building a stock picking engine that actually generated outstanding returns, steadily beating the S&P 500 for over a year on a weekly basis.
Once the engine was ready, I started sending out a weekly newsletter to a handful of subscribers from June 2021 to July 2022, with stock picks for the next week, in conjunction with a simple risk management play, and demonstrated the engine’s prowess through multiple market crises, with steady out-of-sample outperformance available to the newsletter subscribers. In the meantime, I added an engineering co-founder and a marketing maven to assist with the rollout.
And then, for a variety of reasons (mostly to do with the perceived lack of PMF), we pulled the plug and shut down the newsletter service.
