Pros
Smart coworkers, genuinely. The tech stack wasn't embarrassing — gold star.
Cons
There was no coherent ML strategy, or it changed every quarter, which is functionally the same thing. You'd get deep into a project, build something real, and then leadership would shift focus and it quietly gets shelved with no explanation. Feedback cycles showed up either way too late or not at all. Requests came in from people who clearly didn't understand what they were asking for, and there was no real structure for having that conversation. The misalignment between what the business said it wanted and what it actually resourced was constant and exhausting.