Hey everyone,
I just got rejected after interviewing for a Senior Full Stack Engineer role and wanted to share my experience because it’s been sitting with me.
During the technical interview, things actually went really well at first. The hiring manager seemed genuinely impressed — he complimented my experience and how clearly I explained past projects, architectural decisions, and real-world problems I’ve solved in production.
Then we got to the final part: a coding/debugging exercise.
They gave me an AI-generated codebase for a restaurant menu system and asked me to find three bugs. The setup included five separate files, so I had to install the repo, go through each file, understand the overall flow (menu data handling, helper functions, validation, agent logic, and some LLM-related pieces using the OpenAI SDK), and then identify the issues.
Because of the time pressure, I focused on understanding how everything connected. I managed to find one bug — a function returning data with an incorrect type cast instead of properly serializing it — but I ran out of time before I could identify the other two.
What made it more frustrating afterward is that all three bugs were actually located in the same file, which I had been reviewing the whole time. The other four files were essentially there to add complexity, but they didn’t contain any of the issues.
I understand that for senior roles there’s an expectation to quickly navigate unfamiliar code and spot problems under pressure. That part makes sense. But at the same time, it feels like this kind of exercise can outweigh everything else — including experience, system design thinking, and real-world impact — based on performance in a very narrow, high-pressure scenario.
I’m curious if others have been running into similar interview formats lately. Is this kind of “multi-file, AI-generated code, find the bugs” exercise becoming more common for senior roles?
Would appreciate hearing other experiences.
