LFeatured
AIOpenAINode.js
LLM APIs in prod: cost, limits, and guardrails that stick
What we actually measure before shipping OpenAI-style features: tokens, retries, user-visible failures, and when to say no in the UI.
Jafrix System9 min read
Occasional writeups on things we hit in the field: infra, mobile, scheduling, whatever broke that week.