Team Topologies Assessment
A first-90-days assessment from Team Topologies: what kind of team is this, where are its boundaries, and which interaction mode (collaboration, X-as-a-service, facilitating) each neighbour needs.
🌱 A rough idea, planted recently. Expect rough edges — it will grow.
From chapter 5 of Domain-Driven Platform Engineering (Ajay Chankramath and Eamonn Ryan), on Team Topologies and platform as a product. The chapter’s assessments give me a repeatable way to read a team I’ve just joined: what kind of team it is, where its boundaries are, and what each relationship with another team should be.
The three questions
- The team. Which type is it, and how heavy is its cognitive load? (Team Topologies’ four types: stream-aligned, platform, enabling, complicated-subsystem.) A stream-aligned team carrying too much infrastructure is the usual smell.
- The boundaries. Does the team own a whole slice of the domain, or does work need hand-offs to ship? Boundaries should follow the domain (bounded contexts), not tools or the org chart.
- The interactions. For each neighbouring team, which mode fits right now, and for how long?
- Collaboration: working closely for a while to discover something new. Costly, so time-box it.
- X-as-a-service: consume a capability through a clear interface (an API, a golden path; see Golden Paths). The goal state for most platform relationships.
- Facilitating: one team helps another build a skill, then steps away (an enabling team, or a lead coaching a team into new territory).
Platform as a product
The platform team treats other teams as customers: discovery, a roadmap, and adoption that’s measured, not mandated. Success is what it enables, not what it provides (The Platform Relevance Gap, Internal Developer Platform).
Where I’d use it
- First 90 days in a new role: map my team, its neighbours and the current interaction modes. Then agree with the product manager which relationships should move, e.g. from collaboration to X-as-a-service once a contract is stable.
- A team changing shape: a front-end-heavy team taking on backend services needs facilitating for a while (backend skills, on-call, service ownership), then X-as-a-service from the platform, so the new cognitive load doesn’t land all at once.
- #team-topologies
- #platform-as-a-product
- #interaction-modes
- #cognitive-load
- #leadership
- #domain-driven-design
↩ Linked from (0)
Nothing links here yet — it’s a fresh shoot.
↪ Links to (3)
Related by topic
- Seed
Lessons from a Staff Engineer's AI AMA
Platform engineering - Seed
A Safe Path for AI Agents in a Large Codebase
Platform engineering - Seed
Agent Skills and Harness Engineering
Platform engineering - Seed
Brownfield Agentic Engineering
Platform engineering