The 30 second answer
The reported loop has five rounds, and none of them is classic algorithms. There's a multi-part string-parsing screen, an integration round where you extend an existing codebase with HTTP calls and response parsing, a bug squash round in unfamiliar code, an AI-assisted coding round, and finally an experience and goals deep dive. So I'd practise parsing, HTTP and JSON, debugging, testing and clearly explaining my trade-offs.
According to the candidate's report, the loop had five rounds, and the experience round was scheduled only after the four technical rounds were cleared.
- Technical screen. A practical, multi-part coding problem built around parsing a given string. Each part added requirements, and the grading focused on correctness, code structure and edge cases. Prepare by writing small parsers that you extend step by step without rewriting earlier parts, and by keeping functions small so a new requirement touches one place.
- Integration. Implement functionality inside an existing codebase, mostly by making HTTP requests and parsing responses, without breaking existing behaviour, then test against the provided cases. Make HTTP calls and JSON parsing in your interview language automatic before the day, including error status codes and missing fields.
- Bug squash. Find and fix a bug in an unfamiliar codebase. The candidate found and explained the root cause but ran out of time on the fix, and recommended an IDE with a working debugger and an environment set up before the call.
- AI-assisted programming. A multi-part task where AI tools were allowed. The candidate explained why each prompt was written and what output was expected, reviewed the generated code, and wrote and ran tests.
- Experience and goals. An informal deep dive into a technically complex project: ownership and individual contribution, challenges and trade-offs, impact, collaboration and mentoring, and motivation for joining Stripe, with follow-ups on technical decisions and measured impact.