Coinbase Interview Questions — Page 17

Coinbase·Software Engineer·Technical Phone Screen

Jun 2026

Coinbase backend interview focused on extending a recipe management system, which sounds chill until you realize they want you to reason through indexing tradeoffs on the fly. Two feature additions plus a design discussion, all in one round.

You're given a recipe management system with basic CRUD operations backed by a hash map. Add a search function that does case-insensitive substring matching across recipe names and returns all matching recipes.
Add a listRecipes function that returns all recipes sorted by either name or size. Walk through how you'd implement it and whether you'd maintain a sorted structure or sort on demand.

“My first instinct was to just lowercase everything on each scan and call it done.”

View Post

Coinbase·Software Engineer·Technical Phone Screen

Jun 2026

Coinbase backend interview focused on extending an existing recipe management system, which sounds deceptively simple but has some layered design decisions underneath. One question, pretty contained scope, but the follow-up implications are what they're really testing.

Given a recipe management system with existing CRUD and search operations, how would you add a user registration function that tracks which user IDs exist in the system?

“My first instinct was to overcomplicate it.”

View Post

Coinbase·Software Engineer·Technical Phone Screen

Jun 2026

Coinbase software engineer interview focused heavily on probabilistic systems, specifically weighted random selection for NFT minting. The follow-ups kept stacking and by the third one I was just trying to stay afloat.

Design a weighted NFT minter where the constructor takes an array of rarity weights and a mint() function returns a random index proportional to its weight, targeting O(log k) per call using prefix sums and binary search.
How would you support an update(i, delta) operation that changes a weight while keeping mint() at O(log k)? Propose and justify a data structure.
Implement a mintWithoutReplacement(t) function that draws t distinct NFTs according to current weights, updating state after each draw.
How would you make the randomness reproducible via seeding, and what property-based tests would you write to validate the distribution and cover edge cases like zero weights, extremely large weights, all-equal weights, and k=1?

“The prefix sum part clicked pretty fast.”

View Post

Coinbase·Frontend Engineer·Technical Phone Screen

Jun 2026

Coinbase frontend round where they gave me a partially built React blog app and asked me to wire up a form for adding new posts. Pretty focused exercise, more about clean state management than anything algorithmic.

You're given a React app that displays a list of blog posts loaded from a local JSON file. Add a textarea that lets users write and submit a new post, appending it to state so the list updates immediately without a page reload.

“The core of it was keeping the textarea as a controlled input and making sure the list re-rendered from the same state slice rather than some parallel array.”

View Post

Coinbase·Software Engineer·Online Assessment (OA)

Jun 2026

Coinbase software engineer interview, looks like an online assessment style problem centered on NFT tooling. One question, pretty well-scoped, but the constraint about not using retry-based sampling is where most people probably trip up.

Build an NFT metadata generator that takes a collection config and a seed, then outputs N unique trait combinations as JSON. If N exceeds the total possible combinations, return an error. Generation must be deterministic and you cannot use a retry-until-unique approach.

“The seed-based determinism part is straightforward enough, but the 'no repeated sampling' constraint is what makes this interesting.”

View Post

Coinbase·Software Engineer·Onsite - System Design / Architecture

Jun 2026

Coinbase system design round, one big question about pagination that branched into like six different sub-problems. Felt like I was holding on for dear life by the end.

Design a query pagination system. Walk through offset-based vs cursor/keyset pagination, how you handle consistency when rows are inserted or deleted mid-session, deep-page performance, total count queries, ties in sort order, and pagination across multiple shards.

“This started as a pretty normal pagination question and I thought I had it, then they kept pulling on threads.”

View Post

Coinbase·Software Engineer·Online Assessment (OA)

Jun 2026

Coinbase OA for a software engineering role had an AI-assisted coding round where you had to implement a jump boundary function for a Flappy Bird game. The platform rendered the map live which was a nice touch. Passed with 16 out of 20 tests and still moved forward somehow.

Implement a jump boundary function for a Flappy Bird game, where the platform renders the map visually as you work.

“The AI assistant they gave you was pretty useless, barely helped with edge cases.”

View Post

Coinbase·Frontend Engineer·Take-home Assignment

Jun 2026

Coinbase sent over a take-home for a frontend role, basically a full mini-app with a reusable list component, API fetching, reordering controls, tests, accessibility, the works. More involved than I expected for a take-home but it was a pretty clear spec at least.

Build a list component app in React, Vue, or Angular that fetches items from an API on mount and on demand, handles loading and error states, and prevents stale responses from overwriting newer state when requests overlap or the component unmounts.
Implement per-item up/down reordering controls that swap an item with its neighbor, with the first item's up and last item's down disabled. Use stable keys to preserve item identity across renders.
Write unit and component tests covering rendering, data fetching, reordering behavior, and edge cases like empty state and boundary items.
Structure the app with clear separation between data fetching logic and presentation components, and ensure the UI is accessible (keyboard navigable, proper aria-labels, focus management) and responsive.

“The abort/cleanup part is where I spent most of my time.”

View Post

Coinbase·Software Engineer·Hiring Manager Screen

Jun 2026

Behavioral round with the hiring manager for a software engineering role at Coinbase. Heavy focus on AI and LLM project experience, which I wasn't fully expecting from what I thought would be a more standard HM chat. Come prepared to go deep on actual projects, not just vibes.

Walk me through an AI or LLM project you led or contributed to significantly. How did you frame the problem, choose your approach, and measure success?
What trade-offs did you make in your AI project, for example between open-source and proprietary models, or between fine-tuning and prompt engineering?
Describe the team structure on that project and specifically what you personally owned versus what others handled.
If you could continue working on that AI project, what would you do next and why?
Where do you see LLM technology heading over the next couple of years, and how does that shape how you think about building on top of it today?
Why Coinbase, and why this role specifically?
What's the biggest area where you're actively trying to grow right now?

“This was the bulk of the conversation.”

View Post

Coinbase·Software Engineer·Technical Phone Screen

Jun 2026

Coinbase software engineer interview with two algorithmic parts back to back. Started straightforward and then got progressively weirder with the dependency graph twist. Came out of it unsure if I nailed the pruning discussion or just rambled.

Given a sequence of N blocks each with a positive integer value, select a subset to maximize total value such that no two adjacent blocks can both be selected. Implement an efficient solution and analyze its complexity.
Now add dependency constraints to the block selection problem: a directed edge from u to v means block v can only be selected if block u is also selected. The no-adjacent-blocks rule still applies. The dependencies form a DAG. Solve via backtracking with pruning, and discuss when DP over topological order is sufficient versus when full backtracking is necessary.

“Classic house robber variant.”

View Post