Cutting LLM Spend Without Cutting Quality
August 11, 2026
Most LLM cost problems are visibility problems. Per-feature token accounting, caching, model routing, and context discipline — with evals proving quality held.
Read moreTutorials
Field notes on RAG, evals, agents, and LLM cost.
When to Take LangChain Out of the Stack
July 20, 2026
A framework decision is an engineering decision, not a loyalty test. The signals that LangChain is costing more than it saves — and how to exit safely where it is.
Read moreLangGraph Checkpointing in Practice: Durable Agents Without Ceremony
June 22, 2026
Checkpointing is the LangGraph feature that separates production agents from demo loops: resumable state, human-in-the-loop pauses, and replayable failures.
Read moreBootstrapping an Eval Suite From Your First 50 Production Traces
May 18, 2026
You do not need a benchmark or a labeling team to start evaluating an LLM app. Fifty real traces, honest labels, and a runnable harness beat both.
Read moreChunking Is a Retrieval Decision, Not a Preprocessing Step
April 14, 2026
Fixed-size chunking is where most RAG quality problems start. How to choose chunk boundaries from your corpus and your queries — and how to measure the difference.
Read more