Field notes & essays
In-depth, honest articles on custom software, CRM automation, HubSpot, integrations, and the judgement calls in between — written from real work, not summaries.
'Agentic AI' is the phrase of the year, and it's doing a lot of work covering for systems that range from genuinely useful to quietly reckless. Here's a grounded look at where letting AI take actions makes sense — and where it doesn't yet.
An LLM feature that delights ten test users can quietly become slow and ruinously expensive at ten thousand. The economics of AI in production are an engineering problem, and mostly a solved one — if you design for them.
There's enormous pressure to 'add AI' to everything right now. Some of it is genuinely transformative. A lot of it is an expensive way to do worse what a simple rule already did well. Here's how to tell them apart.
Most teams ship AI features on gut feel — it looked good in a few tests, so it went live. Then a prompt tweak silently breaks something and nobody notices. Evaluation is what turns AI from a gamble into engineering.
Everyone can wire an LLM to a vector database in an afternoon. The gap between that demo and a system people trust is almost entirely in retrieval — and retrieval is where most RAG projects quietly fail.
5 articles in AI Engineering