// Systems optimization and life under real constraints
Commercial platforms fail. Support doesn't respond. Bugs accumulate. Meanwhile I'm out here trying to preserve work that matters while also fixing the dryer and hosting Thanksgiving.
"The donkey carries the load regardless of whether the path is paved."
So I'm building infrastructure I control. I'm learning RAG implementation because it's on the critical path. I'm gonna optimize what I can with what I have.
Bash, Perl, Grafana, monitoring, log analysis.
CS50P → AWS certs → Docker → MLOps/Infra.
[2024.11.23]
You don't always get to pick your deadlines. Sometimes a failure on an LLM platform picks one for you. Sometimes it's the dryer timer module. Sometimes it's both on the same weekend.
This is what optimization under constraint actually looks like: it's functional, it ain't perfect.
This is a blog about doing technical work in the real world. You won't find Medium articles with idealized setups here. No bland little half-baked tutorials. Just me, doing my best, with what I have.
This is systems work with:
This donkey wears a hard hat because the work still needs doing, and safety matters even when you're improvising.