Insights
Insights.
Practical notes on building software, modernizing legacy systems and putting AI to work, from the people who do the work.
How many test cases does an AI feature need before launch?
A dozen good examples prove very little. How sample size sets the precision of an AI score, and how to build an evaluation set that catches real failures.Read the articleWhy a fully booked engineering team ships slower
Queueing theory explains why keeping every developer busy stretches lead times, and how planning capacity and limiting work in progress speeds delivery.Read the articleWhy AI agents break on long tasks, and how to design ones that don't
Small error rates compound over many steps. The arithmetic behind unreliable agents, and the two design levers that make them dependable in production.Read the article
Have something to improve?