well, it's not per-engineer. instead, the managers get team-wide metrics so the idea is that they know what the quality of their teams code is, and they can recommend trainings / meetings on how to improve on X part of their AI workflow.