Team Topologies: Organizing Business and Technology Teams for Fast Flow
Over the years, my focus has shifted from Data/AI (Architecture) Design to (Data/AI) Organization Design. The reason: even the best technology is worthless if people don’t work together with the tools.
You can’t build a house with just one craftsman and his toolbox. You need a team of different craftsmen: bricklayers, carpenters, electricians, tilers, plumbers, and many more. They all bring their tools AND their skills with them!
“A fool with a tool is still a fool.”
AI won’t change that. An AI engineer won’t build you a complete AI system. On the contrary, specializations are becoming more differentiated and diverse: prompt engineer, LLMOps engineer, AI architect, data engineer, DataOps engineer, AI compliance manager, etc.
In the end, it’s the team that counts. Or rather, the team of teams. Because often it’s not individual AI craftsmen but several teams: LLM team, MLOps team, data engineering team, data platform team, product teams, and many more.
One book that has helped me, and therefore our customers, to structure their (team of) teams is Team Topologies.
The key message of the book: teams should be stream aligned, i.e. aligned to a specific value chain of the company: They should take end-to-end responsibility!
Other teams can support the stream aligned teams as enabling teams, platform teams or special component teams.
The book also describes a graphic notation for team design, which we also used in some of our workshops in addition to our canvas.
Do you know the book and do you also use the notation? What’s your take on it?
Read the full book review on the Datentreiber blog:
Responses