Articles for people who
ship the stack
Original rewrites on React, Node.js, TypeScript and AI — practical notes from the same engineering practice behind our operator software. Article bodies are in English.
Tagged: production-ai
Moving to Claude Opus 5.5: Five Breaking Changes Behind the Model ID
A ship-gate for switching to claude-opus-5-5: always-on thinking, forced tool_choice, model-bound thinking blocks, computer-use tools and a silent progress stream.
10 min readWhy Agents Pass Every Offline Eval and Still Fail in Production
Six drift modes quietly age an eval set. A four-axis trace score and a weekly cluster-and-promote loop keep CI honest against what live traffic actually does.
9 min readDebug the Pipeline, Not the LLM: A Production Mental Model for RAG Failures
When a RAG bot answers wrong, the LLM is rarely the cause. Trace failures through ingestion, retrieval, generation, evaluation and the loop to find the layer.
10 min read
About these articles
Request a 24h estimate
Need the same stack in a production operator layer? Send the brief — estimate within 24 hours.