Shopify has a Bring Your Own Everything (BYOE) model for coding interviews. That means you bring your own IDE with your preferred AI coding assistants configured. Your coding environment should be set up before the interview.
I applied online and, after the recruiter screen, went through the first round—an AI-enabled coding interview. The recruiter was responsive and informed me of the result the next day: rejected.
Since AI is allowed, it's much harder, as an interviewee, to evaluate what the expectations were and what went wrong.
Sharing my experience with the community.
The question was a standard Shopify-tagged LeetCode problem.
Over the 70-minute interview, I was able to complete the two iterations given to me. I explained my approach and created a prompt for the AI to generate code on top of a boilerplate HTTP server. I then reviewed the generated code, made any necessary adjustments, and explained the implementation to the interviewer.
When asked what else I would add, I talked about global exception handling, proper HTTP 400 and 500 responses, telemetry, business and infrastructure metrics, and API versioning. I also mentioned that, instead of using an in-memory store for cached URLs, I would use an external cache or database to make the solution scalable.
The interviewer was responsive throughout, and I didn't feel that communication was the problem.
The entire interview was recorded, and at the end I had to submit a pull request. I believe the solution is reviewed after the interview, possibly with AI assistance, so the outcome may not be based solely on the interviewer's assessment.
My takeaway from the rejection is to start with a minimal boilerplate, talk less and code more, and implement as many production-ready features as possible. In an AI-enabled interview, simply talking about improvements may no longer be enough when you are evaluated on your usage of AI during the interview.
