OpenAI · Primly Community

interviewing at openai for PM in 3 weeks, can people share recent loops?

quietquit_quincy · 5 replies

got a recruiter screen coming up and want to understand the full process before I'm in it.

I've seen older posts (2022-2023) saying the PM loop was lighter on product design and heavier on research discussion, but I don't know if that's still true given how much the company has changed.

anyone who's been through the PM process in the last 6-8 months: what rounds did you see? was there a product critique, an estimation question, a strategy case? did they ask you to present anything?

also curious how much of the behavioral is specifically about AI experience. I've built AI-adjacent features but haven't worked on model development directly. is that a problem?

would really appreciate any real data points, even if you didn't get the offer.

5 replies

jordan_pm

went through it for a product role in the research division about 5 months ago. loop was: recruiter screen, PM phone screen (behavioral + product sense, 45 min), then onsite with 4 rounds: product design, data/metrics, strategy, behavioral. no take-home. they did NOT ask me to present anything in advance.

the product design question was specifically about an AI product. hard to dodge that. think through what makes AI features different to design for.

apm_aisha

super useful. did the behavioral panel feel separate from the product rounds or did interviewers mix in situational stuff throughout?

jordan_pm

mixed throughout, honestly. the product design interviewer asked a behavioral in the last 5 minutes. just be ready for it anywhere.

director_dee

the AI experience question is worth taking seriously. you don't need to have trained models, but you should be able to talk intelligently about how AI products fail, where user trust breaks down, what responsible rollout looks like. that's the actual bar.

ux_uma

went through UXR (not PM) recently but the behavioral component sounds similar. they asked me specifically about navigating situations where user research conflicted with what the model could do. be prepared for AI-specific constraint scenarios.