Microsoft System Design Interview: Design Google Maps
Watch a mock system design interview where a college graduate with internship experience at Uber and Facebook designs Google Maps with a Microsoft software engineer. This full-length interview covers everything from requirements gathering to database sharding — a common system design question at top tech companies.
The interview walks through designing a navigation system at massive scale: 300 million monthly active users and approximately 200 million route queries per day. You'll see how to break the globe into geographic grid squares, represent road networks as weighted graphs, implement routing with Dijkstra's algorithm, and separate static map data from dynamic traffic data. The discussion covers caching trade-offs for route queries, cross-block routing strategies, ETA calculation using real-time traffic as edge weights, and database sharding for distributed map storage. The interviewer's feedback: "for a college grad, you went pretty far."
0:00 Introduction and interview setup
1:15 Problem statement: Design Google Maps
2:30 Clarifying requirements (routing, ETA, transport modes, traffic)
4:00 Scale estimation: 300M MAU and 200M daily route queries
6:10 Geographic grid system and graph representation of roads
9:00 Address resolution and routing service design
11:30 Dijkstra's algorithm for shortest path routing
13:45 Cross-block routing across geographic boundaries
15:30 Caching trade-offs for route queries
17:00 ETA service and real-time traffic as edge weights
18:30 Database design: static map data vs dynamic traffic data
19:45 Sharding strategies and rebalancing
20:30 Interviewer feedback and performance review