"Tell me about a time you worked closely with a Data Science or Machine Learning team on an ambiguous project at scale — for example, building a training dataset, feature pipeline, or experiment dataset used in Hive or a production workflow. What was the cross-functional challenge, how did you handle disagreements or unclear requirements, and what was the outcome?"
This question tests whether you can collaborate effectively with partners who have different goals, vocabulary, and success metrics. At Meta, Data Engineers often work with DS, ML Engineers, and product teams to define data contracts, improve data quality, and unblock decision-making or model development without direct authority over the other team.
Interviewers want to see whether you can create clarity when requirements are still forming, influence technical and non-technical stakeholders, and take ownership beyond simply delivering a table or pipeline. They are also looking for how you handle tension: conflicting metric definitions, shifting scope, data quality issues, or pressure to move fast.
A strong answer is a specific STAR story with real stakeholders, a concrete data asset or decision, and measurable impact. The best responses show how you aligned on definitions, used data to resolve disagreement, made trade-offs explicit, and improved the collaboration model for future work.