Ask HN: What's lacking in your internal build and deploy pipeline

I work on a Kubernetes based infrastructure platform and am interested to know what issues do service developers face.

We can constrain the conversation to stateless services and batch workloads, keeping DBs outside, since they require a more hand managed solution from what I see.

How do you think of: 1. Tenancy 2. Multi-cluster strategy 3. Monitoring 4. Shared responsibility between service devs and infra providers?

It would be interesting to hear from all sides - service devs, infra devs, sres and devops engineers.