I was at a Series C startup that suffered from this. The CTO was an SRE person at heart so wanted to build out our own infrastructure that could run on any plat

I was at a Series C startup that suffered from this. The CTO was an SRE person at heart so wanted to build out our own infrastructure that could run on any platform. While this was a great idea in theory (basically everything was kubernetes and terraform) in practice it caused us to hire a ton of SRE people that were not contributing to the features that actually made us money. Not one customer cared that our entire service ran on AWS. I think a lot of companies suffer from premature optimization. I myself even fell victim to it in one of the projects I worked on for that company by rewriting a large portion of the code instead of hacking the new feature in. It is a hard problem to solve for. How much tech debt do I take on now to deliver customer value. I guess that is something that just comes with experience though. I have noticed that has I gain more experience I am not enamored by the new shiny framework as much anymore and focus more of my effort on delivering whatever feature is needed with the smallest change possible. Yeah this creates tech debt, but at least it keeps the lights on especially when every service seems to get rebuilt in 5-10 years anyways.