Exams / AIF-C01 / ML lifecycle, MLOps and metrics / Question 4
AIF-C01 practice question: ML lifecycle, MLOps and metrics (4 of 4)
Fundamentals of AI and ML A free question from our original AIF-C01 practice exam. Try it, then open the answer to see why each option is right or wrong.
An insurer's data scientists retrain models by hand on their laptops and copy files to production when they remember. The CTO wants to adopt MLOps. Which TWO practices are part of MLOps? (Choose two.)
Choose two answers.
- AA. Version the data, code and models so results can be reproduced
- BB. Train each model once and never change it after release
- CC. Automate training, testing and deployment in a repeatable pipeline
- DD. Let each scientist choose their own manual release process
- EE. Skip testing so that new models reach production faster
Show the answer and explanation
Answers: A, C
- A. Right: Correct. Version control of ML assets lets teams reproduce a result and roll back to an earlier model.
- B. Wrong: Data changes over time, so models degrade. MLOps expects monitoring and retraining, not a one-off release.
- C. Right: Correct. Automation makes each stage repeatable and consistent, and supports continuous training and delivery.
- D. Wrong: Ad hoc manual releases are the problem MLOps is meant to solve. They are hard to audit or repeat.
- E. Wrong: MLOps speeds up delivery through automation, including automated testing. It does not remove quality checks.
Explanation
MLOps applies DevOps ideas to ML: version control for data, code and models; automated, repeatable pipelines; continuous monitoring of data and model quality; and retraining when performance drops. The goal is faster, more reliable delivery with an audit trail.
Study
Keep practising
Answer questions with a score at the end, or practise just this topic with the explanation after each one. No sign-up needed.