Zero-downtime rollouts on Kubernetes
A practical checklist for shipping services without taking users offline — probes, PDB, and traffic shifting.
- AK
- DM
An observability budget that scales
How we keep metrics, logs, and traces useful without drowning the bill — sampling, SLOs, and ownership.
- DM
- AK
DNS that doesn’t lie at 3 a.m.
Records, TTLs and cutovers that keep migrations calm — plus the checks we run before touching production names.
- AK
RAG without the hype
Retrieval-augmented generation that stays operable — chunking, evals, and failure modes that actually show up in production.
- MS
- AK
An eval harness before the demo
Minimal evaluation for LLM features — golden sets, regression gates, and what to measure when accuracy is fuzzy.
- MS
- AK