Google System Design Questions

Design a Billing-Token Rate Limiter for AI Agents

Design a distributed rate limiter that protects APIs by limiting requests (or consumed units such as LLM billing tokens or storage quota) per user, API key, tenant, or IP. It must support configurable rules (e.g., 100…

Also reported as: Design A Drive-Style Usage Quota Limiter · Design a large-scale distributed rate limiter

Design a Large-Scale Polite Web Crawler

Design a distributed web crawler that starts from seed URLs and crawls about 10 billion pages (or all images reachable from given root URLs), with periodic recrawls. It must avoid fetching the same URL twice, obey robots.txt and…

Design a Personalized Weekly Deals Service

Design a recommendation system from scratch, for example recommending posts, games, local sports teams or weekly deals to users. Cover candidate generation (collaborative filtering, content-based, social signals, location),…

Match Tagged Problems to Skilled Programmers at Scale

Given a set of problems tagged with required skills and a set of programmers with skill profiles and capacity, design a system that assigns problems to suitably skilled programmers at scale. Start with the algorithm (model it as…

Design a Task Scheduler for Opaque Long-Running GPU Jobs ("Design Sora")

Design an asynchronous text-to-video generation service like Sora. Users submit a prompt; generation takes minutes on scarce, expensive, failure-prone GPUs. Users can see job status and progress, cancel jobs, and download the…

Design a Security Monitoring Framework

Design a security monitoring framework for cloud infrastructure: collect security telemetry (audit logs, network flows, host events, IAM changes), detect threats with rules and anomaly detection, generate and prioritize alerts,…

Design an Online Coding Judge Platform

Design an online coding platform like LeetCode or Codeforces. Users browse problems, write code in the browser in many languages, submit it, and get a verdict (accepted, wrong answer, TLE, runtime error) after it runs against…

Design Calendar Event Conflict Handling

Design a calendar service like Google Calendar. Users create, update and delete events (one-time and recurring), invite attendees who can accept or decline, view day/week/month/agenda views across time zones, share calendars, get…

Design an Embeddable Streaming Chat Interface

Design a ChatGPT-style conversational AI product. Users send a message and receive a streamed response from a model; conversations persist and can be resumed on any device; the service must stay highly available despite limited,…

Visualize More Than One Hundred Million Map Points

Design an interactive map that visualizes more than 100 million data points. At broad zoom levels it shows aggregated heat-map or cluster information; at close zoom it shows individual pins, with smooth panning and zooming.…

Design an Enterprise AI Adoption Agent for a Global Company

Design an AI agent that helps a large global company adopt AI across teams: it answers questions using internal data, calls internal tools on behalf of employees, and respects security, data-residency and access-control…

Design a pub-sub replay system

Design a pub/sub system that supports replay: subscribers can rewind and re-consume messages from a past point in time or offset (e.g., after a bug fix), in addition to normal real-time delivery. Discuss the durable log storage,…

How to host many domains on one IP?

You have one Linux server with a single public IPv4 address, and three businesses (coffee shop, butcher, auto repair) each want their own website on their own domain. Explain how to host all three: DNS A records pointing to the…

Design street-view image ingestion and storage system

Design an ingestion and storage system for Street View imagery: vehicles capture huge volumes of high-resolution images with GPS and pose metadata, upload them (in bulk from drives or over networks), and the system processes…

Design a global real-time notification system

Design a multi-channel notification service (push, email, SMS, in-app) that other services call to notify users. It must accept sudden bursts of millions of notifications with low acknowledgement latency, persist them durably,…

Design nearest-available-taxi lookup with updates

Design the core of a ride-hailing platform like Uber. Riders request a ride from their location; the system finds nearby available drivers, dispatches the request, and tracks the trip in real time. Drivers report their location…

Design large-scale near-duplicate video detection

Design a large-scale near-duplicate video detection system (e.g., to catch re-uploads of copyrighted content that have been cropped, re-encoded or overlaid). For each new upload, quickly find videos in a corpus of billions that…

Design a Twitter hashtag metrics aggregator

Design a system that continuously computes the top K items from a very large event stream: for example the K most-viewed URLs, the K services with the most error logs in the last 5 minutes, the top 100 most active users in a…

Design a global restaurant menu update system

Design a menu update system for a global restaurant chain. Headquarters and regional managers publish menu and price changes; each restaurant has several kinds of menu-display devices (digital boards, kiosks, POS, tablets) that…

Design lexicographic range query word store

Design a persistent storage structure for a large set of strings that supports adding and deleting words and lexicographic range queries: given [L, R], return all words w with L <= w <= R (optionally paginated or with a limit).…