Posts from 2026
31 posts
The Hidden Cost of Trusting AI Without Checking Its Work
AI is good at sounding right. That is exactly why it is dangerous. Why you should treat AI output as a starting point, not a source of truth.
Your AI Meeting Notes Are Losing Context
AI meeting summaries compress conversations into bullet points, but downstream agents need the full transcript to produce useful output.
Using FFmpeg to Let Claude Code See a Client's Bug
How to convert a client's screen recording into sequential image frames so Claude Code can visually analyze the bug being demonstrated.
How I Wrote a Python Script to Parse Google Timeline Data Instead of Just Installing a Mileage App
How I reverse-engineered Google Maps Timeline JSON data with a Python script to extract months of business mileage instead of installing a tracking app.
AI Web Research Has a Fact-Checking Problem
Web fetch tools can return confidently wrong prices, features, and specs alongside accurate content. Here's the verification step that caught it.
Let's Talk
No sales pitch. No commitment. Just direct contact with the engineer.
Folks We've Helped
