Kevin Badinger
WorkLabsProjectsBlogResumeOne-PagerContact

Writing

Thoughts on building software that lasts, AI that actually helps, and lessons from two decades of production code.

2026

Mar 31

The Demo Read the Clean PDF. Production Sent a Fax.

Reading the right characters is only part of document extraction. The value still has to reach the right field in the final record.

Mar 17

An Accuracy Score Cannot Reconstruct One Wrong Decision

A disputed credentialing decision shows what a case history can explain that an aggregate accuracy score cannot.

Mar 3

Before Rewriting a Legacy System, Read Its Exceptions

An old payment retry check shows why a rewrite needs to investigate exceptions before deciding which behavior to keep.

Feb 21

The Market Is Adversarial. Your Prompt Isn't.

Can an LLM actually trade? Not 'help you trade' or 'make you feel like you're trading smarter'—actually generate real returns as a bot. The $9–11 billion AI trading platform market really wants the answer to be yes. The evidence says something different.

Feb 19

The AI Coding Illusion - Why AI-Generated Demos Fall Apart in Production

The demos look great. AI writes code faster than you can type and the result looks like a week's work in 45 minutes. The problem isn't the demo. The problem is thinking that's where software development ends.

Page 4 of 12
← Previous123456789101112Next →
Twitter/X • LinkedIn • GitHub
Last updated April 2026