Question bank
4,310 interview questions, answered.
Reference answers, what the interviewer is really testing, how it is graded, and the follow-ups that come next.
Easy 190Medium 2,498Hard 982
Filters · onDifficulty, topic, company
2,498 questions
- 865What is your least favorite product among the ones you use, and how would you improve it?MediumProduct·
Google, Meta - 866PermutationsMediumCoding·
Adobe, Apple - 867How would you leverage discovery features like search to acquire users for TikTok Shop?MediumProduct·
TikTok - 868How would you leverage smartphone smelling technology?MediumProduct·
Google - 869Lexicographic Grid Travel.MediumCoding·
Databricks - 870Should LinkedIn acquire Discord? Why or why not?MediumProduct·
LinkedIn - 871Installs on LinkedIn's iOS app are down, what do you do?Medium·
LinkedIn - 872If LinkedIn messages are down, how would you identify the cause?MediumDataProduct·
LinkedIn - 873What could be the drivers for LinkedIn's performance in different markets?MediumDataTechnical·
LinkedIn - 874You're a PM at LinkedIn. Your data science team reports that 64% of outbound messages receive no response. What do you do?MediumDataBehavioral·
Google - 875What are three endpoints on Discord?MediumTechnical·
Discord, Solenis - 876List four endpoints on Tinder.Medium·
Tinder - 877Bias-variance tradeoff: practical rules of thumbMediumAI ML
- 878BPE vs SentencePiece vs WordPiece tokenizationMediumAI ML
- 879Dense vs sparse (BM25) vs hybrid retrievalMediumAI ML
- 880Diagnosing bias vs variance by experimentMediumAI ML
- 881Document chunking strategies for RAGMediumAI ML
- 882Grid search vs random search vs Bayesian optimizationMediumAI ML
- 883Handling severe class imbalance in productionMediumAI ML
- 884LoRA vs full fine-tuningMediumAI ML
- 885Model Context Protocol (MCP)MediumAI ML
- 886RAGAS: why it's popular and where it breaksMediumAI ML
- 887The ReAct frameworkMediumAI ML
- 888Top-k, top-p (nucleus) and temperature samplingMediumAI ML