← Back to Directory

Coinbase interview notes, page 4 of 21

Large Enterprises

Most of these Coinbase write-ups come from Software Engineers (122), then Data Scientists (41) and Product Managers (25). The most reported rounds are the technical phone screen (112), the system design round (31), the product sense round (25) and the online assessment (11).

210 interview notes · updated Jul 2026

Coinbase·Software Engineer·Onsite - Coding / Algorithms

Apr 2026
Coinbase software engineer interview with a coding round built around an in-memory order management system for crypto trades. The problem looked manageable at first glance but the constraints pushed you toward thinking carefully about data structure choices rather than just getting something working.
  • Design and implement an in-memory crypto order management system that parses commands from stdin. Orders have an id, symbol, side, quantity, and state (LIVE, PAUSED, CANCELLED). Support CREATE, PAUSE, RESUME, CANCEL, GET, and COUNT commands, printing output for queries and ERROR for invalid operations or bad state transitions. The system needs to handle up to 200,000 commands efficiently.

“The problem itself is not crazy hard but the scale hint is what tripped me up at first.”

View Post

Coinbase·Data Scientist·Technical Phone Screen

Apr 2026
This was a deep-dive case question for a DS role at Coinbase, basically one long structured scenario about diagnosing a revenue drop. No fluff, no warmup. The whole session was just this one beast of a question.
  • A key revenue metric on Coinbase's dashboard drops noticeably week-over-week. Walk through your full investigation: how do you confirm the drop is real, decompose the drivers, segment the data, diagnose the funnel, audit recent changes, separate external from internal causes, run hypothesis tests, and set up monitoring going forward.

“This question is enormous.”

View Post

Coinbase·Data Scientist·Technical Phone Screen

Apr 2026
A pretty intense DS case round at Coinbase that was basically a full e-commerce revenue diagnosis problem. Lots of moving parts and they expected you to go deep on causality, not just surface-level decomposition.
  • Weekly revenue dropped 20% versus the 4-week moving average. Walk through how you'd decompose that delta into traffic, conversion rate, and average order value, and then break those down further into price effects, mix effects, discounts, and post-purchase deductions like returns and cancellations. Show a waterfall with approximate percentage-point contributions.
  • What are the fastest checks you'd run within 24 hours to falsify or confirm your hypotheses about what's driving the revenue drop?
  • How would you design a causal model to separate the impact of the new free-shipping threshold from the competitor's sitewide sale and from normal seasonality? What control groups, pre-trend checks, and robustness tests would you use?
  • What experiments and fixes would you prioritize to address the revenue decline, and what success metrics would you track for each?

“This is the core of the whole case and I spent way too long setting up the framework before getting to actual numbers.”

View Post

Coinbase·Software Engineer·Onsite - System Design / Architecture

Apr 2026
System design round at Coinbase for a software engineering role. The whole thing was one big question about designing a crypto order placement system, and it went deep fast. Walked out unsure if I'd covered enough of the failure handling side.
  • Design an order placement system for a crypto trading platform that routes buy and sell orders to multiple third-party exchanges, each with different protocols and reliability profiles.
  • Your request to a venue times out after you've already sent the order. What state do you record, and how do you safely retry without duplicating the order?
  • Walk through how a user's funds should be represented and reserved when an order is placed, so you can reconstruct the full history of every balance change for an audit.
  • A venue ACK is lost but the order actually filled. Reserved funds are stuck and the user tries to resubmit. How does reconciliation detect and fix this without double-charging or double-filling?
  • How do you cancel an order that was split across two venues when one leg has partially filled and the other is still open? What does the user see during that window?
  • How do you guarantee an execution report is applied exactly once to the ledger when the venue might redeliver messages and your consumer could crash mid-update?
  • How would you extend the design to handle stop-loss or conditional orders, where the trigger depends on live market data rather than being submitted immediately?

“Big open-ended one.”

View Post

Coinbase·Software Engineer·Hiring Manager Screen

Apr 2026
Interviewed for an EM role at Coinbase. Just one question from what I can tell, but it's a meaty one that I probably underprepared for.
  • How do you measure the performance of an engineering manager?

“I went straight to output metrics and kind of forgot about the people side for a bit.”

View Post

Coinbase·Software Engineer·Onsite - System Design / Architecture

Apr 2026
Coinbase system design round, basically a deep dive into building an in-memory database from scratch. They kept layering on follow-ups until I was pretty deep in the weeds on transactions and TTL.
  • Design an in-memory key-value store with basic get, set, and delete operations. Walk through your data structure choices and the complexity of each operation.
  • How would you add TTL (time-to-expiration) support to the key-value store? What are the tradeoffs between lazy expiration and eager background cleanup?
  • Add transaction support to the database: BEGIN, COMMIT, and ROLLBACK, including nested transactions. How do you handle isolation and rollback semantics?
  • How would you implement snapshot and restore functionality? What are the space and time tradeoffs of different approaches?
  • How would you support secondary lookups, like scanning by a field value or key prefix? What data structures would you add and what does that cost you?

“Started with a hash map, which is the obvious answer, and they just nodded and waited.”

View Post

Coinbase·Data Scientist·Technical Phone Screen

Apr 2026
Coinbase data scientist interview with a deep A/B testing case built around a real KYC drop during a volatile BTC period. The question was layered enough that I kept second-guessing whether I was overcomplicating it or not going far enough.
  • KYC completion rate on the mobile onboarding funnel dropped 8% week-over-week during a period of high BTC volatility. Design an A/B test to improve completion without increasing fraud risk. Cover your randomization unit, primary and guardrail metrics, how you'd control for market shocks, sample size and duration, sequential monitoring, country-level compliance variation, and a pre-registration plan.

“This thing is basically six questions stapled together.”

View Post

Coinbase·Software Engineer·Technical Phone Screen

Apr 2026
Coinbase software engineer interview with a follow-up coding question building on a key-value store from a previous round. Pretty focused on data structures and knowing when to reach for the right one.
  • You have a key-value store where every operation on a key increments a counter for that key. Given that setup, return the top N keys ranked by their operation count, highest first.

“This is a continuation of a prior question so you're already sitting with a working key-value store implementation.”

View Post

Coinbase·Software Engineer·Technical Phone Screen

May 2026
Coinbase software engineer interview with a pretty meaty system design question centered on a food delivery platform. One long problem broken into three parts, each building on the last. Not the hardest thing I've ever seen but there's enough surface area to trip you up if you're not careful about scope.
  • Given a set of restaurants (each with a menu, item prices, and a geographic location) and a stream of incoming orders, implement a way to find: (a) the restaurant that can fulfill a given set of menu items at the lowest total price, and (b) the restaurant geographically closest to the user.
  • Given a list of orders and their item prices, compute the following over a specified time window: total revenue, number of orders, and average order value.
  • Over the same time window, return the top-K orders ranked by total price, and the top-K menu items ranked by transaction volume.

“I jumped straight into the geo part because it felt more interesting, which was probably the wrong move.”

View Post

Coinbase·Software Engineer·Onsite - System Design / Architecture

May 2026
Coinbase system design round, one big question that basically ate the whole session. The scope was huge and I don't think I covered everything cleanly, but it was a solid technical conversation.
  • Walk through the entire lifecycle of a user uploading a photo: from the moment they hit submit in the browser, through auth, storage choices, validation, async processing, CDN delivery, and everything in between.

“This thing just kept expanding.”

View Post