2026
an archive of posts from this year
| Jul 21, 2026 | Prompt caching in production: where the savings disappear |
|---|---|
| May 28, 2026 | Why agent memory degrades in production |
| Apr 08, 2026 | Context engineering as a production discipline |
| Jan 25, 2026 | Why LLM evaluation metrics look stable but customers are unhappy |
| Jan 18, 2026 | Context limits degrade routing quality faster than generation |
| Jan 11, 2026 | RAG fails upstream |
| Jan 04, 2026 | Embeddings for intent classification: architecture trade-offs |