ML for Trading in the Age of AI Agents
Stefan Jansen
Founder, Applied AI · Author, ML4T
This course is popular
3 people enrolled last week.
Work through an ETF strategy from idea to a realistic trading simulation
Learn how the decisions in a trading research project fit together, then work through them with me on a prepared ETF example.
In four Python notebooks, you'll inspect market data, construct features and future-return labels, train a model on successive time periods, and turn its forecasts into a portfolio. You'll compare the same backtest before and after commissions and spread, and learn how to interpret prediction quality, uncertainty, turnover, and drawdown.
I'll demonstrate a coding agent implementing part of the pipeline from a written brief. We'll also inspect a recorded research-agent forecast to see how specialists gather evidence, use tools, and combine estimates. These examples introduce both coding agents and agent engineering in a research workflow.
If you've read Machine Learning for Trading or worked through individual tutorials, this session connects the research decisions to the code and results. You'll leave with notebooks, an agent brief, slides, a recording, and a workflow you can adapt to your next research question. The live session includes guided notebook work and Q&A.
What you’ll learn
Connect a trading hypothesis to data, model forecasts, and portfolio results, with reusable notebooks for your own research.
Follow a momentum hypothesis through feature definitions, the forecast horizon, and the portfolio rule.
Inspect data coverage and liquidity eligibility, then work through momentum, volatility, RSI, and forward-return labels.
Run LightGBM over successive periods, purge overlapping labels, and interpret forecast rankings and uncertainty.
Backtest a monthly portfolio of the ten highest-ranked ETFs, then compare commissions, spread, turnover, and drawdown.
Watch a coding agent implement part of the pipeline, then inspect a recorded forecast from a team of research agents.
Identify the further research, risk controls, execution checks, and monitoring needed after a backtest.
Workshop agenda
The ML for Trading research workflow (25 min)
Follow the decisions from market hypothesis through data, modeling, validation, and portfolio evaluation. Presenter-led introduction.
Specify a momentum hypothesis (10 min)
Connect an economic explanation to measurable features, a forecast horizon, and conditions under which the hypothesis may fail. Presenter-led.
Inspect the ETF data (35 min)
Guided work in notebook 01: inspect histories, liquidity eligibility, returns, and correlations in the provided ETF data.
A coding agent implements part of the pipeline (20 min)
Presenter-led demonstration using a written brief and checkpoints for labels and evaluation. A 10-minute break follows.
Construct features and labels (30 min)
Guided work in notebook 02: calculate momentum, volatility, RSI, and liquidity features, then check future-return labels.
Train and evaluate the model (45 min)
Guided work in notebook 03: train LightGBM over successive periods and interpret its forecasts and uncertainty. A 10-minute break follows.
Backtest the portfolio with trading costs (40 min)
Guided work in notebook 04: turn model forecasts into a monthly ETF portfolio and compare performance, trades, and turnover with and without costs.
Inspect a research-agent forecast (20 min)
Presenter-led walkthrough of a recorded trace: specialist evidence, disagreements, aggregation, and the final forecast.
From backtest to live-trading requirements, and Q&A (25 min)
Discuss further testing, execution, risk controls, and monitoring. Review your options for further study and answer your questions.
Learn directly from Stefan
Stefan Jansen
Builds production ML and agent systems, then teaches the operating method.
Who this workshop is for
Readers of Machine Learning for Trading who want guided practice connecting the research process, Python code, and portfolio results.
Developers and data scientists who know Python and want an introduction to financial research with a complete ETF example.
Practitioners new to ML for trading who want to see coding and research agents used within a concrete research workflow.
What's included
Live sessions
Learn directly from Stefan Jansen in a real-time, interactive format.
Attendee repository
Four research notebooks, ETF data, and the research trace.
Coding-agent brief
Reuse the written specification shown in the live demo.
Slides and recording
Revisit the notebook walkthroughs and explanations.
Maven Guarantee
Your purchase is backed by the Maven Guarantee.
Frequently asked questions
Maven for Teams
Reimbursement
Get your company to pay
Everything L&D needs: email template, receipts, and certificate of completion.
Get reimbursedTeam discount
Learn with your teammates
Save 20%+ when 2 or more teammates enroll in the same cohort.
Save 20%+ with a teamPrivate cohort
Run a cohort for your org
A dedicated cohort with a custom schedule and curriculum, tailored to your team.
Book a private cohort$299
USD
4 days left to enroll
10am–2:30pm EDT