How full stack developer interviews work
Full stack developer interviews test depth across the entire development stack. Expect a coding round, a frontend-specific round covering JavaScript and component architecture, a backend round covering API design and server-side logic, a system design round, and behavioral questions. Some companies replace algorithm questions with a take-home project.
The balance between frontend and backend depth varies by role. Read the job description carefully: which technologies are listed first and what does the team actually ship? Tailor your preparation accordingly rather than splitting effort equally across all areas.
JavaScript and frontend questions
"Explain the difference between == and === in JavaScript." == performs type coercion before comparison, so 0 == false is true. === checks both value and type without coercion, so 0 === false is false. Use === in all cases unless you specifically need type coercion. "How does the browser event loop work and why does it matter for UI performance?" JavaScript is single-threaded. The event loop processes the call stack, then microtask queue (Promises), then macrotask queue (setTimeout, I/O). Understanding this explains why long synchronous operations block UI updates and why heavy computation should be moved to Web Workers.
Backend and API questions
"Design a REST API for a task management application." Define resources (tasks, users, projects), HTTP methods for each operation, status codes, authentication approach, and pagination for list endpoints. Show that you think about the API from the consumer's perspective, not just the implementation side.
"How would you handle database transactions where two operations must both succeed or both fail?" Use a database transaction: begin, execute both operations, commit on success, rollback on any failure. In ORMs like Prisma, transactions are wrapped in a callback. The key is ensuring partial state is never persisted if either operation fails.
Database questions
"When would you use SQL versus NoSQL?" SQL suits structured data with complex relationships and strong consistency requirements. NoSQL suits flexible schemas, very high write throughput, and data shapes that map poorly to relational tables. The decision should be about access pattern fit, not general preference. "Explain database indexing tradeoffs." Indexes speed up reads via an ordered structure the database can binary search. The cost: indexes use disk space and every write must update all relevant indexes. Over-indexing a write-heavy table is a common performance mistake.
System design questions
"Design a URL shortener like bit.ly." Cover: API design (POST to create, GET to redirect), database schema, unique short code generation (hash or base62 sequential ID), caching hot URLs in Redis, and CDN for static redirects at scale. "How would you make a slow page load faster?" Work through the stack: network (CDN, HTTP/2, compression), frontend (bundle splitting, lazy loading, image optimisation), API (reduce requests, add caching), database (query optimisation, N+1 elimination). Measure first, then fix the biggest bottleneck.
Behavioral questions
"Tell me about a technical decision you made that you later regretted." Show that you identified the problem, understood root cause, drove the correction, and changed how you make decisions as a result. "How do you stay current across both frontend and backend?" Be specific: newsletters, open source projects, side projects. Show your learning is deliberate and that you have a way to filter signal from noise. Full stack developers who cannot keep up with either side quickly become a liability on modern teams.