Practice in the browser
Each company tends to favour a particular flavour of coding question. Pick a company to practise the problems that commonly appear in its interviews, solve each one in a real editor, and run it against the test cases. These are community-reported favourites, not official lists, and interview content varies by team and level.
Google coding rounds lean algorithmic and favour clean, well-reasoned solutions over raw speed. Graphs, dynamic programming, and a willingness to discuss complexity out loud go a long way.
Meta runs fast, medium-weighted rounds where finishing two problems cleanly matters. Arrays, strings, and tree or graph traversal show up constantly, and pace counts.
Amazon pairs practical data-structure problems with its leadership-principle behavioural bar. Graph and BFS questions, stacks, and clean two-pointer work are common, and clear reasoning is rewarded.
Microsoft rounds tend toward the classics: linked lists, strings, and trees, with an emphasis on correctness and edge cases rather than exotic tricks.
Apple coding rounds favour solid fundamentals and careful implementation. Expect arrays, strings, and a design-flavoured data-structure question, with attention to detail over speed.
Netflix keeps its loops small and senior, so each problem carries weight. Practical algorithmic questions with a real discussion of trade-offs are the norm.
Uber leans on graphs and simulation, fitting for a logistics company, alongside the usual medium array and string problems. Clear modelling of the problem matters.
Airbnb is known for grounded, applied problems, often parsing or string-heavy, sometimes wrapped in a small real-world scenario. Readable, correct code is valued.
Stripe is the outlier: its rounds favour practical, correctness-focused problems over textbook algorithms, often with real input parsing and careful edge-case handling rather than clever asymptotics.
ByteDance rounds run harder and more algorithmic than most, with a real appetite for dynamic programming and graph problems at speed.
LinkedIn favours trees, graphs, and design-leaning questions, often with a follow-up that pushes on scale or an alternative data structure.
Snap keeps its coding rounds in familiar medium territory: arrays, strings, and the occasional graph, with clean implementation expected.