GuideAdvanced

Hiring and Evaluating Engineers in the AI Era Guide

When AI writes a large share of the code, the signals you used to hire with stop separating candidates: the take-home comes back perfect, live coding from memory measures something else, and in a remote interview you cannot tell whether the answer came from the person or from a hidden backing track. And when everyone uses agents, counting closed tickets no longer tells you who is performing. This guide teaches you to hire and evaluate engineers in that world: which profile to look for, which signal still predicts performance, how to design a structured process with a clear AI policy at each stage, how to detect fraud without becoming paranoid, how to calibrate the bar with a deadline on top of you, what to do about juniors, how to make the process fair, how to onboard the people who join, and how to evaluate performance by impact, judgment and review, not by volume. You learn it as the Director of Engineering at **Nómada Software**, a LATAM nearshore consultancy with 120 engineers that needs to hire 30 in six months. This year, 3 of the 24 people hired left before 90 days, one of them was not the person who was interviewed, and the performance cycle rates almost everyone as "exceeds expectations". Module by module you build the answer: a diagnosis, the role scorecards, the interview process with its AI policy, an integrity protocol, the bar and its calibration, a plan for juniors, fairness and onboarding, and a performance evaluation framework. There is no code to write. The math is done in tables and by hand, and the studies come with their source and date.