Blog
Measuring developer productivity, one question at a time
Oren Keinan, Aidealy's founder, on what to measure in an engineering team, what to pair it with, and what to leave alone. Each article comes with its video, the full transcript and its sources.
Do DORA metrics still work with AI-generated code?
10-min video
DORA (originally DevOps Research and Assessment) metrics still work when half your code is AI-written, but only in pairs: a speed number is safe to read only next to a stability number.
Do tokens or lines of code measure developer productivity?
10-min video
Measuring developer productivity keeps failing when you count one activity, from lines of code in 1982 to the AI token counts on Amazon's leaderboard in 2026.
Which developer productivity metrics should you never use?
10-min video
Developer productivity metrics that count a single activity, from lines of code to AI token counts, get cheated, and when they decide reviews or layoffs, the wrong people get hurt.
How to measure developer productivity when AI writes code
10-min video
To measure developer productivity when AI coding tools like Claude Code and Cursor write much of the code, put every number for speed or amount next to the number that drops when the work is bad.
From Oren's LinkedIn newsletter
See what your R&D is really doing.
Tell us what you're trying to figure out, and we'll get you set up to answer it on your own R&D: your engineers, your AI agents, and the code they ship together.
Talk to us