[Seeking advice / lessons learned on CI/CD platform products] Hey everyone, we are having a VERY fragmented CI/CD approach at our company, with over 85 product teams and 3500 repos. We have some alignment across some repos (using project imports in gitlab), but mostly fragmented. Product teams don't have a lot of time to invest into CI/CD, so we are on a rather basic level and don't have any central leverage to across our stack because the deployment methods are very different. To increase efficiency, reduce cognitive load of teams, and to drive overarching features for everyone (e.g., canary deploys, security features, ...), we want to build a platform product to offer to our product teams. ❓ What are the experiences of folks with a similar, highly fragmented starting point? What did work? What didn't work? How have you approached the problem?

My initial thoughts: • Of your 85 product teams, start where you can make the most difference for the least amount of effort - don't try to fix everyone's problems all at once. There's probably a team or two that have an easier stack, simpler pipelines, etc. and are willing to work with your team for some quick wins. • Work with them on ADRs that identify (and attempt to solve) the pain points with an MVP platform product, make sure you and them are agreeing on the tools, architecture, and overall solution • Form a steering committee of champions across your org, folks who are interesting and willing to change, to help tell the story of what you are doing and provide insightful feedback (and not just "this is great!" or "I hate this!" but actual feedback) • From this point, rather than trying to create a super long reply to your query in Slack, I have a handful of personal blog posts on platform engineering from similar experiences here: ◦ 1 https://wahlnetwork.com/2023/10/05/platform-engineering-is-product-design-at-scale/ ◦ 2 https://wahlnetwork.com/2023/09/08/things-ive-learned-leading-a-platform-engineering-team/ I hope this helps in some way!