Dataford
Top 50
Topic roadmap
Updated weekly · Last refresh Sep 12

Top 50 ML Ranking Interview Questions

The most frequently asked ML Ranking questions across all roles and companies, ranked by real interview frequency. Updated daily.

50questions
~7htotal time
59companies covered
Track your progressSign up free to work through all 50 questions and resume where you left off.
Start practicing free →
1
ML RankingStart here. 50 questions · ~400 min
Design Facebook Feed RecommenderHard
Recently asked

Design the end-to-end ML system for Facebook Feed recommendation, from retrieval and ranking to serving, evaluation, monitoring, and failure handling.

ML RankingRetrievalRecommendation SystemsMetaMeta Platforms
Design a Short-Video Retention RecommenderHard
Recently asked

Design a short-video recommendation system that balances immediate engagement with long-term retention in a personalized feed.

ML RankingRetrievalRecommendation SystemsYouTubeByteDance
Design Instagram Stories RecommenderHard
Recently asked

Design Instagram Stories recommendation at Meta scale using retrieval, ranking, re-ranking, and robust monitoring under a 120ms p99 budget.

ML RankingRetrievalRecommendation SystemsMeta
Design a Behavior-Based RecommenderHard

Design a recommendation system that uses user behavior to retrieve, rank, and re-rank items at scale.

ML RankingFeature StoreRecommendation SystemsPanasonicOpera SolutionsAurora Payments
Design Sparksoft Personalized RecommendationsHard

Design Sparksoft's personalized recommendation system, including training, multi-stage serving, evaluation, monitoring, and production failure handling.

ML RankingFeature StoreRecommendation SystemsMetaOpera SolutionsBest Buy
Choose Retrieval vs Ranking StackHard

Design a large-scale shopping recommender and decide when two-tower retrieval beats a traditional ranking stack.

ML RankingTwo-Tower ModelsRecommendation SystemsSalesforceAvisoNorthwestern Mutual
Design a Digital Banking Offer RankerHard

Design an ML system that ranks banking offers for online customers using retrieval, ranking, and re-ranking.

ML RankingRecommendation SystemsRegions FinancialApplied Data FinanceCapgemini FSSBU
Design Safe Content Recommendation FilteringHard
Recently asked

Design a safety-aware recommendation stack that prevents harmful content from being recommended at 350M DAU and 2.2M peak QPS.

ML RankingFeature StoreRecommendation SystemsPluralsightMetaTikTok
More ML Ranking questions with a free account

Sign up to see every question

Create a free account to unlock this list and practice real interview questions.

Get my prep plan
The finish line: interview-readyComplete all 50 questions to finish this plan.