This message was deleted.

I think the most important thing is to well understand what do you want to improve with platform engineering: speed up deployment? allow developers consuming infra autonomously? … What is important for me is to well identify the major pains and implement solutions in the right direction. Depending on companies organisations speaking with users (aka developers ) is not always easy.

This is spot on @jean-philippe Foures

Fully agree. Platforms are just like any other software - we need to perform business & system analysis before implementing anything!

Absolutely, building a platform no one wants/needs doesn’t solve any problems. If the new fancy tool provides value then great, otherwise it’s not worth it.

True 🙂 This is a big difference with standard IT project management where focus is on specifications, Platform Engineering is fully based on the user need. Classical example of our Platform Engineer routine : support (developer requests), documentation of our internal infrastructure products catalog, new feature design, ...

Some hints on what we need to take care of besides writing k8s operators: https://khalasa.com/2024/01/how-to-approach-internal-developer-platform-delivery/