MLOps interview questions and answers

You track the histogram of scores or class rates over time and compare to a baseline.

TTL expires stale keys so you do not serve last month's session counts as if they were live.

Continuous training automatically retrains on new data when triggers fire.

Precompute when the candidate set is large and latency is tight, then serve a lookup.

Data paths, feature lists, hyperparameters, seeds, and compute settings should be declarative YAML or Hydra configs.

Automated comparison of canary versus baseline metrics with statistical tests, not a human squinting at graphs.

If you wait too long to fill a batch, every request pays that wait, inflating p99.

Compiled runtimes can cut latency and GPU memory versus eager Python.

A producer adds a new enum or changes amount from int to float and serving fills NaNs.

Label delay is the time until the true outcome is known, such as 30-day default.

CI runs tests on code and sometimes data validation.

Docker freezes dependencies so the model runs the same in laptop, CI, and cloud.

The distribution of model outputs changes over time.

So experiments are reproducible and audits can explain a model.

New model scores live traffic without affecting users.

Training code unit tests, data schema checks, and a smoke train on a tiny sample.

A small slice of traffic proves metrics before full rollout.

Training/serving skew from duplicated feature code.