About
AI in dealflow often feels like a black box — powerful but opaque. In this hands-on session, we’ll build a simple ML model on a startup dataset and use SHAP to make predictions explainable. Participants will see how founder and company attributes influence outcomes, and learn to separate meaningful signals from noise. Drawing on recent founder-stage-fit research, we’ll show how explainable AI can support IC workflows, reduce bias, and provide insights investors can actually trust and replicate.
Details
🧭
15:30 - 16:20
🌟
🎤
Workshop
Speakers
%201.png)
Michael Mansfeld
Investment Manager | Head of Data & Product
Michael Mansfeld
Investment Manager | Head of Data & Product
,
HPI Ventures
Michael joined HPI Ventures in 2023 and is currently Principal & Head of Product, where he leads the development of the firm’s sourcing and dealflow platform. His focus is on building data-driven systems that enhance how investment opportunities are identified, evaluated, and supported. Before HPI Ventures, Michael co-founded neopolis, a legal tech startup building a “GitHub for lawyers” that enabled collaborative drafting and version control for legal documents. He also served as a Digital Policy Advisor to the German Parliament, working at the intersection of technology, regulation, and innovation. Alongside his investment work, Michael is is an external PhD student at TU Munich, researching explainable machine learning methods for venture success forecasting and a Research Affiliate at MIT DesignX, where he explores how AI can power venture design and decision-making.
Moderators
No items found.
Keynotes
No items found.