Notes from the field.

Jun 18, 2026

6 min read

How do you stop LLM hallucinations in document extraction?

The fix isn't a smarter model — it's refusing to trust any model's output until it's proven against the source document. Here's the validation pattern we use in production.

May 27, 2026

5 min read

Why shouldn't your data pipelines run on a timer?

Polling every minute for new files feels safe. It's actually the single easiest way to overload an orchestration platform for no benefit. Here's what to do instead.

May 4, 2026

6 min read

Rules, ML, or LLMs — how do you choose for document automation?

The right answer is usually 'all three, tiered by cost' — not picking one technology and forcing every document through it. Here's the framework.

Apr 12, 2026

7 min read

What does agentic AI actually look like inside a real enterprise?

Not one model doing everything — specialized agents, each owning one decision, checked by a separate validator agent before anything ships. Three proofs of concept, one pattern.

Mar 20, 2026

5 min read

What are SEO, AEO, and GEO — and why should a business owner care?

SEO gets you found on Google. AEO gets you quoted by Alexa and voice search. GEO gets you cited by ChatGPT and other AI assistants. Here's what actually changes each one.