Most enterprise AI initiatives do not fail on raw model intelligence. They break on the mundane, structural realities of production environments: latency budgets, deterministic fallback paths, and auditable…

Just spent the afternoon untangling a “helpful” ORM layer that was executing N+1 queries under the hood. Sometimes raw SQL is the only abstraction you need.

Talked to a few people this week who admitted to pasting confidential docs into a public chat tool because the sanctioned internal one is slower and worse. This…

Ran intervals on the track this morning wearing two watches, one with track-mode distance correction and one without. 0.72% distance error on the one with it, several times…

The pgvectorscale benchmark everyone is citing this year: 471 QPS at 99% recall on 50 million 1536-dim vectors, competitive with dedicated vector databases. If you’re already running Postgres,…

Re-read the Meadows bathtub explanation for the third time this year. Stock and flow is such a simple idea and I still catch myself forgetting it mid-meeting. Someone…

Donella Meadows’ list of places to intervene in a system has one line that has been sitting with me for weeks: people know intuitively where leverage points are,…

The AI coding and drafting tools got good enough this year that the bottleneck moved. It used to be: can we build it. Now it’s: should we, and…

Every RAG demo looks good. Ten hand-picked questions, confident answers, a happy stakeholder in the room. Three weeks into production a prompt update breaks tool calling, retrieval starts…

SR-26-2 replaced SR-11-7 this year, the first update to US bank model risk guidance in fifteen years. The interesting part isn’t the new rule, it’s what forced it.…