MLOps Basics — ML Technical Interview Questions & Answers

Versioning, monitoring and retraining models in production.

MLOps Basics

  • Simple meaning: MLOps is running ML like software: version data + code + model, test before ship, monitor live, retrain when the world changes.
  • Picture: Last year’s fail predictor on this year’s new exam pattern — accuracy silently dies. Monitoring catch that. A pipeline retrains on new labelled rows.
  • Basics to name: Git, model registry, CI, drift checks, rollback to last good model. You don’t need Kubernetes in a fresher answer — name the problems.
  • Trap: “we’ll retrain by hand someday”. Trap: monitoring only server CPU, not prediction quality.

Technical Interview Tip: Treat the model like an app: versions, logs, retrain when data shifts.

Code Examples for MLOps Basics

MLOps Basics — working example

# PCA-style intuition: project 2D → 1D by a direction vector
points = [(1.0, 1.0), (2.0, 2.0), (3.0, 2.8)]
# unit direction along y = x
ux, uy = 0.7071, 0.7071
proj = [x * ux + y * uy for x, y in points]
print([round(v, 3) for v in proj])

# Tiny ensemble: average two model votes
votes = [1, 0, 1]
v2 = [1, 1, 1]
ens = [1 if a + b >= 2 else 0 for a, b in zip(v1, v2)]
print(ens)

MLOps Basics — dim reduction compresses features; ensembles combine models; deep nets/RL/NLP are specialized families — practice the core math idea first.

How to Explain MLOps Basics to an Interviewer

What is data drift?

Live X no longer looks like train

Name three MLOps ideas.

Version the model Monitor Retrain