Interview questions / AI / Machine Learning Engineer
AI / Machine Learning Engineer interview questions
These are questions AI / Machine Learning Engineer candidates often hear, with a short note on what a strong answer covers. Reading them helps; saying your answers out loud helps more.
Walk me through taking a model from notebook to production.
Cover packaging, serving, monitoring, retraining and rollback.
How do you know a deployed model is getting worse?
Mention data drift, prediction drift, delayed labels and alerts.
How would you build a feature that answers questions about company documents with an LLM?
Discuss retrieval, chunking, evaluation, cost and handling wrong answers.
Explain the bias-variance trade-off with an example.
Keep it concrete and link it to a model choice you made.
How do you evaluate a model when labels are scarce?
Mention small labelled sets, human review, proxy metrics and active learning.
Tell me about a model that didn't work. What did you learn?
Be honest about the failure and specific about the lesson.