Shopify interview questions

Tech / SWE · 20 questions · last reported May 2026
Report an interview →

Browse 20 Shopify interview questions for software engineering and software development interview preparation. Questions in this collection are tagged for full-time, internship roles. Explore the topics and stages below and filter the collection to plan your practice.

Interview stages in this collection
Phone screen, Onsite, Online assessment (OA)
Practice topics
behavioral (6), system-design (4), coding (3), pair-programming (3), algorithms (2), arrays (2), career-motivation (2), class-design (2)

Questions come from candidate contributions and online research. Reports reflect past experiences and do not guarantee what Shopify will ask. Locked question details require points. Public examples below are free to read.

Read 5 free Shopify interview question examples

Browse all Shopify questions

20 questions

Free Shopify interview question examples

Read the full problem statements and use the links to revisit or share a specific example.

LRU Cache

Medium · Phone screen · Onsite · Full-time

The pair programming question was to implement an LRU cache. I had to support the usual get and put behavior, and evict the least recently used item once the cache hit capacity. The interviewer cared about whether the code was clean and actually usable, not just whether I knew the textbook answer. I used a hash map with a linked list to track ordering, though an OrderedDict in Python would work too. After the implementation we spent some time testing the behavior and making sure edge cases were handled.

Datasource Class

Medium · Phone screen · Onsite · Full-time

This one was more practical than a straight LeetCode question. I was asked to implement a Datasource class that behaves kind of like a tiny Pandas DataFrame. The example was something like creating a Datasource from a list, chaining map and filter calls, and then calling collect at the end. The main thing was making the chaining work cleanly while keeping the data transformations understandable. It felt like they were testing API design, class design, and whether I could make a small abstraction feel natural to use.

Project Ownership

Medium · Onsite · Phone screen · Full-time

They asked me to pick a project I'd owned end-to-end and walk them through it in detail. I had to draw out the architecture, explain the components, and justify why things were designed the way they were. The interviewer kept asking "why" at every layer, not in an aggressive way, but they clearly wanted to see if I actually understood the system versus just having worked on it. They were really interested in how I handled tradeoffs, like where I cut scope versus where I pushed back, and what I'd do differently in hindsight. It felt like the most real engineering conversation of the whole loop.

CS Motivation

Medium · Onsite · Phone screen · Full-time

So the round was split into two parts. First they asked me why I chose to go into computer science, which felt pretty personal compared to most behavioral questions I've gotten. Then we jumped into a 45-minute coding problem involving arrays and lookups, somewhere around medium difficulty. They wanted me to talk through my approach as I coded, not just silently solve it. I had to optimize for edge cases and walk through the time complexity at the end. The interviewer seemed to care a lot about whether I could explain my reasoning clearly while writing code at the same time.

Webhook Delivery

Hard · Onsite · Phone screen · Full-time

The system design round was probably the hardest part of the whole loop. They wanted me to design a webhook delivery system from scratch, specifically focused on how to handle retries when deliveries fail. I went with exponential backoff and the interviewer pushed me on edge cases like what happens when a downstream consumer goes offline for hours at a time. We spent a good chunk of time talking about idempotency and whether ordering guarantees actually matter for webhooks in practice. They also wanted to know how I'd handle permanently failing endpoints, so I brought up dead-letter queues and alerting. It felt less like a textbook design question and more like a real problem they'd actually dealt with on the platform.