Stripe Interview Process & Rounds

Stripe's loop is built from its own round types rather than LeetCode. Reports describe an online assessment on some loops, one technical screen, and a virtual onsite. The assessment, reported by new-grad, intern and experienced candidates alike, is a single long multi-part problem on HackerRank of about 60 minutes. The technical screen is one multi-part coding problem, most often shipping cost, payment-to-invoice matching or business data verification. A full-time onsite is usually four or five rounds: a programming exercise; an Integration round inside a cloned repository, where Bike Map is the most-reported question in the catalog; a Bug Squash round that asks you to fix failing tests in an unfamiliar library, most often a modified Mako template engine; a system design round; and a behavioral or hiring-manager round. Intern loops report two onsite rounds, and some L2 loops report no system design. From spring 2026, several reports say the programming round, and on some loops the screen, became an AI coding round using HackerRank's built-in assistant. Most coding-style rounds come in parts that build on each other, and more than one candidate reports passing without finishing the last part.

Key facts

  • 7 distinct round types
  • 70 questions reconstructed from 405 candidate reports
  • Reports span Jan 2025 – Sep 2026
  • Refreshed monthly · last updated September 2026

The Stripe loop, from candidate reports

Stripe's loop is built from its own round types rather than LeetCode. Reports describe an online assessment on some loops, one technical screen, and a virtual onsite. The assessment, reported by new-grad, intern and experienced candidates alike, is a single long multi-part problem on HackerRank of about 60 minutes. The technical screen is one multi-part coding problem, most often shipping cost, payment-to-invoice matching or business data verification. A full-time onsite is usually four or five rounds: a programming exercise; an Integration round inside a cloned repository, where Bike Map is the most-reported question in the catalog; a Bug Squash round that asks you to fix failing tests in an unfamiliar library, most often a modified Mako template engine; a system design round; and a behavioral or hiring-manager round. Intern loops report two onsite rounds, and some L2 loops report no system design. From spring 2026, several reports say the programming round, and on some loops the screen, became an AI coding round using HackerRank's built-in assistant. Most coding-style rounds come in parts that build on each other, and more than one candidate reports passing without finishing the last part.

What does Stripe ask in each round?

Stripe interviews span 7distinct round types, shown below. Counts reflect distinct questions per round across the loops we’ve indexed.

The technical screen and the onsite programming exercise: one business problem in parts that build on each other. Shipping cost, the subscription email scheduler, business data verification and payment-invoice matching lead the reports.
31 questions
Most-reported: Shipping Cost Calculator (36×)
One long multi-part problem on HackerRank, about 60 minutes, reported on new-grad, intern and experienced loops. The Jupyter load balancer and the request routing system are the most reported.
14 questions
Most-reported: Jupyter Load Balancer (8×)
Fix failing tests in a modified open-source library. Mako is the long-running repository, requests, Moshi/Jackson and SnakeYAML appear in other loops, and interviewers weigh the debugging process as well as the fixes.
9 questions
Most-reported: Mako Template Engine (27×)
Design at the API and data-model level. The ledger, a metric counter and a feature flag SDK are the most reported.
7 questions
Most-reported: Ledger System (10×)
Hands-on work inside a cloned repository against real files and HTTP APIs. Bike Map dominates; invoice reconciliation against an API and request replay also recur.
6 questions
Most-reported: Bike Map (46×)
A behavioral round and a hiring-manager round: why Stripe, your proudest project, feedback and conflict, and in 2026 reports, how you use AI at work.
2 questions
Most-reported: Behavioral Round (22×)
A coding problem solved with HackerRank's built-in AI assistant, reported from 2026 in place of the programming exercise and on some screens. The transaction rule parser is the reported question.
1 question
Most-reported: Transaction Rule Parser (9×)
What passing candidates do
  • Writing readable, modular code instead of racing to the last part. One report that ended in an offer credits clean naming, structure and edge cases over completion; on the shipping-cost problem, a candidate whose first two parts used well-named helpers only had to change one helper for Part 3.
  • Working the AI coding round as plan, prompt, review. Candidates who did well formed their own view of the problem before prompting, read the assistant's code closely, and added their own tests. At least one candidate who sketched their own skeleton first was asked why they were not using the AI.
  • Using a real debugger in Bug Squash. Candidates who debugged Mako with print statements report running out of time. A tip from one rejected candidate: compare a failing test with the passing tests next to it, which usually differ only in their input.
  • Unblocking yourself in Integration. The round gives you a repository and documentation, not guidance, and reports say it measures how quickly you find your way around unfamiliar code and choose your next step. One rejection's feedback named exactly that.
  • Designing at the API and data-model level. Reports on the ledger round say it wants API-level detail rather than only a high-level diagram, and it probes idempotency keys and how money is modelled.
Where candidates lose points
  • Choosing Java. Several reports say the multi-part rounds are very hard to finish with Java's boilerplate; one intern repeated 'do not use Java' three times, and one interviewer said the bar would be higher for a Python candidate.
  • Spending the round reading. The problem statements are long: candidates on the email scheduler report that reading time left too little time to debug, and data verification reports call the statement too long to finish reading. On invoice matching, misreading the input format cost candidates the later parts.
  • Bugs in your own test cases. More than one candidate lost many minutes debugging a correct solution against a wrong expected value copied from an earlier test.
  • An environment that will not run. On Mako, a broken IDE debugger cost some candidates the round: one traced it to a Python version the repository did not support and failed, another fixed only one bug.
  • Answering staff-level behavioral questions with execution stories. One staff candidate was told the technical rounds were fine but the behavioral answers did not show staff scope.

FAQ

How many rounds is the Stripe interview?
Stripe's loop spans 7 round types: The technical screen and the onsite programming exercise: one business problem in parts that build on each other. Shipping cost, the subscription email scheduler, business data verification and payment-invoice matching lead the reports., One long multi-part problem on HackerRank, about 60 minutes, reported on new-grad, intern and experienced loops. The Jupyter load balancer and the request routing system are the most reported., Fix failing tests in a modified open-source library. Mako is the long-running repository, requests, Moshi/Jackson and SnakeYAML appear in other loops, and interviewers weigh the debugging process as well as the fixes., Design at the API and data-model level. The ledger, a metric counter and a feature flag SDK are the most reported., Hands-on work inside a cloned repository against real files and HTTP APIs. Bike Map dominates; invoice reconciliation against an API and request replay also recur., A behavioral round and a hiring-manager round: why Stripe, your proudest project, feedback and conflict, and in 2026 reports, how you use AI at work., A coding problem solved with HackerRank's built-in AI assistant, reported from 2026 in place of the programming exercise and on some screens. The transaction rule parser is the reported question..
What is the Stripe interview process?
Stripe's loop is built from its own round types rather than LeetCode. Reports describe an online assessment on some loops, one technical screen, and a virtual onsite. The assessment, reported by new-grad, intern and experienced candidates alike, is a single long multi-part problem on HackerRank of about 60 minutes. The technical screen is one multi-part coding problem, most often shipping cost, payment-to-invoice matching or business data verification. A full-time onsite is usually four or five rounds: a programming exercise; an Integration round inside a cloned repository, where Bike Map is the most-reported question in the catalog; a Bug Squash round that asks you to fix failing tests in an unfamiliar library, most often a modified Mako template engine; a system design round; and a behavioral or hiring-manager round. Intern loops report two onsite rounds, and some L2 loops report no system design. From spring 2026, several reports say the programming round, and on some loops the screen, became an AI coding round using HackerRank's built-in assistant. Most coding-style rounds come in parts that build on each other, and more than one candidate reports passing without finishing the last part.
How fresh is this Stripe interview data?
It's reconstructed from candidate reports spanning Jan 2025 – Sep 2026 and refreshed monthly as new reports come in.

Get the full Stripe catalog

Every question, every candidate-reported follow-up, and what passers actually do. Monthly refresh.

Is this helpful?