Why we built Maven
Every hiring platform bans AI from assessments. We built one that scores how you use it. Here's the thesis, the data, and why we think the entire industry has it backwards.
Field reports, candidate behavior studies, and engineering deep-dives on what makes a great developer in the Copilot era.
Every hiring platform bans AI from assessments. We built one that scores how you use it. Here's the thesis, the data, and why we think the entire industry has it backwards.
A 0-to-100 measure of how effectively a developer works with AI tools. We break down the five behavioral dimensions, how they're weighted, and what separates a 90 from a 50.
After analyzing hundreds of coding sessions, we identified nine distinct patterns in how developers collaborate with AI — from the Architect who plans before prompting to the Speedrunner who iterates in rapid-fire cycles.
We compared assessment scores against 90-day hiring outcomes. The signal that correlated most strongly wasn't code correctness — it was how candidates handled AI suggestions they disagreed with.
A practical guide for hiring managers: how to write assessment briefs that reveal AI collaboration skill, what to look for in reports, and common mistakes that produce false signals.
A technical deep-dive into how Maven captures and analyzes developer behavior: timeline events, session reconstruction, behavioral fingerprinting, and scoring calibration.
New posts land every two weeks. Email blog@maven.dev with "subscribe" to get them in your inbox.