"Tell me about a time you had a conflict with a coworker while working on a machine learning project at scale — for example, around model quality, launch timing, data readiness, or evaluation criteria for a surface like Facebook Feed or Ads ranking. How did you handle it, and what was the outcome?"
At Meta, ML engineers work across research, product, data science, infrastructure, and software engineering. Conflict is normal; what matters is whether you can address disagreement directly, stay focused on the user and business impact, and move the work forward without damaging trust. Interviewers are looking for how you handle tension when there is no clear authority line, incomplete information, or competing priorities.
A strong answer uses one specific example with real stakes, explains both sides of the disagreement, and shows how you used evidence, communication, and ownership to resolve it. The best responses are structured in STAR format, quantify the result, and include what you learned or would do differently next time.