I read the article and it's still not clear to me how they are trying to solve the bottlenecks they talk about

After reading their main page I saw that they are just another provider of CI that claims to be faster. Not sure about others but to me the problem with CI/CD has never been that it's slow. The painful points are setting it up and getting it to work. Once that's done everything after it's not that painful

Do the investors know that "compute layer for code" doesn't mean anything and is total nonsense?

It does mean something to me, but perhaps not as profound as whoever coined the term was hoping!

A "compute layer for code" is called a microprocessor and a lot of companies already make them.

Sounds cool, but their custom setup makes it way too hard to switch back if you ever want to leave.

If they want to win over big teams, they should make it really simple to move your current code over to it without a headache.

I'm building a product in the same space[1] and, in my opinion, the core problem is that small tasks meant to run as workers end up spinning up a full 64GB server instead. My approach is to separate lightweight checks (e.g. "is this commit tagged as version-*?") into workflows that run on Cloudflare Workers. They're fast, cheap, and highly available. The heavier CI/CD work can then run on usual docker contaiers/VMs.

1: https://codeinput.com/products/workflow-automation - The product is still in development. Happy to connect with anyone interested in beta testing or contributing ideas!