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
982 questions
- 313Imagine you are a data scientist for Instagram. How would you balance ads and follower posts and how would you monitor its effectiveness.HardData·
Meta - 314You're a PM at Bing. What would you do to take over 5% of Google's search market share?HardDataProduct·
Microsoft - 315Design a solution for borrowing and lending on Facebook.HardProductSystem design·
Meta - 316How would you bring Waymo to market in five years?HardProduct·
Google - 317How would you build an ad platform for Robinhood?HardProduct·
Robinhood - 318How would you build a tool to authenticate products sold on TikTok Shop?HardProduct·
TikTok - 319Build a Basic Regex ParserHardCoding·
Adobe, Apple - 320How would you build a Facebook product for blood donation?HardProductSystem design·
Meta - 321Build a new feature for natural disaster response.HardProduct·
Meta - 322Build a Google campus on the moon.HardProduct·
Google - 323How would you build Google Wing?HardProduct·
Google, Google X - 324How would you build an object detection system to identify and locate flowers in images?HardAI ML·
Apple - 325How would you build a system on a ticketing website to manage high traffic during simultaneous ticket purchases?HardProduct·
Google, Meta - 326How would you build TinyURL?HardSystem designTechnical·
Google, Meta - 327How would you build a recommendation system for an e-commerce store?HardDataTechnical·
Amazon, Lyft - 328How would you build a monitoring device for elderly people?HardProduct·
Google - 329Design a self-driving car solution for kids.HardProduct·
Google - 330Debug a transformer with 4 failing tests, then train a classifier (OpenAI)HardCodingSystem design·
OpenAI - 331Debug an actual LLM inference step in a 20–30 line Colab skeleton (Anthropic)HardCodingSystem design·
Anthropic - 332Design a centralized ML training and management platformHardCodingSystem design· MEMeta engineer mock
- 333Design a distributed training pipeline for an LLM with fault toleranceHardCodingSystem design·
Anthropic - 334Design a fine-tuning platform (instruction / PEFT / RLHF)HardCodingSystem design·
OpenAI - 335Design a RAG system with evaluationHardCodingSystem design·
OpenAI - 336Design a system for safe deployment of AI models in productionHardCodingSystem design·
Anthropic