Courses
Work with me on machine learning for trading
Two ways to work with me on machine learning for trading: an 8-week live cohort and a focused 1-day workshop.
Taught by Stefan Jansen — author of Machine Learning for Trading (Wiley, 3rd ed. 2026) and maintainer of the 19K-star ML4T open-source ecosystem.
Live cohort · Maven · 9 markets + production discipline · September 7 to October 31, 2026 · $2,200
Machine Learning for Trading: From Research to Production
For people who already run end-to-end ML pipelines and want a research process they can apply to their own work across markets.
Turn a working ML trading pipeline into a repeatable research process: baseline, diagnose, iterate, account for costs, and prepare the work for production review. You work across nine case studies spanning ETFs, crypto perps, NASDAQ microstructure, S&P options, firm characteristics, FX, CME futures, and an equities panel. Your capstone is your own strategy taken through a full research loop, reviewed 1-on-1 with me.
Cohort: 8 × 90-minute live sessions with me, 4 office hours every other week, and a 30-minute 1-on-1 capstone review per student.
Saturday August 1, 2026 · Maven · 10am to 3pm ET · $650
Engineering a Multi-Agent Forecasting System
For people building agentic systems in finance who want the architecture, failure modes, and evaluation discipline in one focused Saturday.
Build a multi-agent forecasting system you can inspect, test, and score. A single live day, alternating lecture and lab. During the day, you build a working repo of teaching notebooks and a local Ollama setup, based on chapter 24 of the book (Bridgewater AIA replication).