Build Multi-Agent Systems You Can Audit
Wed, Jun 24, 2026
3:00 PM UTC (30 minutes)
Virtual (Zoom)
Free to join
Go deeper with a course
Building Multi-Agent Forecasting Systems
Stefan Jansen
Author, ML for Trading (Ch. 24: agents) · Founder, Applied AI
Wed, Jun 24, 2026
3:00 PM UTC (30 minutes)
Virtual (Zoom)
Free to join
Go deeper with a course
Building Multi-Agent Forecasting Systems
Stefan Jansen
Author, ML for Trading (Ch. 24: agents) · Founder, Applied AI
What you'll learn
What makes a forecast scoreable, not just persuasive
Five stages — question, evidence, independent runs, aggregation, scoring — that decide whether your number holds up.
Audit while the agent works, not after
Replayable evidence, independent runs, explicit aggregation — captured as the trace forms, not bolted on at the end.
The result most agent demos hide
Brier, log score, calibration locked before resolution — including when agent + consensus beats either alone.
Why this topic matters
Most AI forecasts collapse under one question: "Why should I trust this number?" Auditability isn't bolted on at the end — it's decided five stages earlier: the question, the evidence, the independent runs, the aggregation rule, and the scoring plan locked before resolution. Build the process; the audit takes care of itself.
You'll learn from
Stefan Jansen
Author, ML for Trading · Founder, Applied AI · Investing since 2013
Stefan is the author of ML for Trading — the book and open-source companion code (19,000+ GitHub stars) that have become a practitioner reference for applying ML to financial markets. The 2026 third edition expands to nine cross-asset case studies, with a foreword by Antonio Gulli, Senior Director, Google.
He maintains the Zipline fork the quant community relies on, and built the six-library stack — data to live — behind the third edition's case studies. Investment partner since 2013, he has built trading platforms and live strategies across asset classes. In 2016 he founded Applied AI, which brings production ML to investment teams and other data-rich verticals. He has taught ML to 110,000+ professionals through DataCamp and General Assembly, incl. at Bloomberg and BlackRock.