How do I approach ML System Design interview questions?

ML System Design questions require understanding of core concepts and practice. PracHub provides solutions with explanations to help you master ml system design interviews.

What difficulty level is this interview question?

This is a medium difficulty ML System Design question, commonly asked during Onsite rounds at Meta.

What role is this question designed for?

This question is commonly asked for Machine Learning Engineer candidates at Meta during technical interviews.

Design a newsfeed dislike model | Meta Interview Question

Quick Overview

This question evaluates a candidate's ability to design and operationalize a machine learning model that predicts the probability of a user disliking a social newsfeed post, testing competencies in label design, training data construction, feature engineering, model selection, offline evaluation, online experimentation, and serving and ranking integration. It is in the ML System Design category for Machine Learning Engineer roles and is commonly asked to assess practical experience with production ML pipelines and metric trade-offs; the problem requires both conceptual understanding of design trade-offs and practical application skills for deployment and handling new signals and offline/online metric discrepancies.

Design a machine learning system for a social newsfeed that predicts the probability that a user will dislike a post.

Assume there is already an existing production model that predicts P(like) for ranking. Describe how you would design the new P(dislike) model, including:

problem definition and label design
training data construction
feature design
model choice
offline evaluation
online experimentation
serving and integration into ranking

Then answer two follow-up questions:

How would you leverage a new feature: whether the post has received at least one like in the last 6 months?
What would you do if offline metrics improve, but the online metric does not improve after deployment or A/B testing?

Quick Overview

Design a machine learning system for a social newsfeed that predicts the probability that a user will dislike a post.

Assume there is already an existing production model that predicts P(like) for ranking. Describe how you would design the new P(dislike) model, including:

problem definition and label design
training data construction
feature design
model choice
offline evaluation
online experimentation
serving and integration into ranking

Then answer two follow-up questions:

How would you leverage a new feature: whether the post has received at least one like in the last 6 months?
What would you do if offline metrics improve, but the online metric does not improve after deployment or A/B testing?

Design a newsfeed dislike model

Quick Overview

Solution

Submit Your Answer to Earn 20XP

Design a newsfeed dislike model

Quick Overview

Solution

Submit Your Answer to Earn 20XP