Team Topologies for ML Teams

Building Effective and Thriving Machine Learning Teams - David Tan & Dave Colls

Topic 5 of 9

Published Jan 27, 2025 · Clip 23:32–31:41 (8:09)

David Colls applies the Team Topologies model to ML teams: stream-aligned teams deliver end to end, platform teams abstract common compute/data concerns, complicated subsystem teams encapsulate specialized ML capabilities like car valuations as self-service APIs, and enabling teams consult on recurring issues like privacy and ethics. David Tan illustrates with a complicated subsystem team exposing car valuation as an API that mobile, web, and email teams could integrate with. They close with Deming's quote that 'a bad system will beat a good person every time,' emphasizing creating systems that help teams build the right thing in a way that's right for people.