RepostedStripe logo

Stripe · Phone screen → Onsite

In process

Stripe New Grad Screens and Onsites: 9 Experiences (Part 2 of 2)

Software engineer · New GradBachelor's degreeSep 2026
Shared by

Stripe new grad screens and onsites, Nov 2025–Feb 2026: 8 reports built on CSV and record parsing, shipping-cost tiers, fraud rules and Bikemap.

At a glance

Coverage: 8 Stripe engineering candidates (entries 16, 18–20 and 23–26): SWE and SDE new grads with no location given, plus one Machine Learning Engineer (25) and one onsite of unconfirmed level (19). Report dates November 2025 to February 2026. Four team-screen reports once numbered here (15, 17, 21 and 22) are covered in the Stripe new grad team-screen reports. A separate SDE report with level not given (S1, September 2026) sits at the end, outside the counts.

In this series

  • Part 1: entries 02–06 and 11–13, the newer timelines from February to October 2026, including two offers.
  • Part 2 (current): entries 16, 18–20 and 23–26 from November 2025 to February 2026, plus the merchant-link report S1.

What came up

Most screens here handed candidates text or CSV records and a growing list of rules. Counts are out of the 8 core reports; one candidate can sit in several rows.

Topic Reports Last seen
Multipart coding 5 of 8 Feb 2026
Record parsing and validation 5 of 8 Jan 2026
Payment, fee and pricing rules 4 of 8 Jan 2026
API and repository integration 2 of 8 Jan 2026
Repository debugging 2 of 8 Jan 2026
ML modeling 1 of 8 Nov 2025

Questions we found

Algorithms / Data Structures

Question Reports Last seen
Shipping Cost Part 1 and Part 2 2 Dec 2025
Project and Filter Structured Records 2 Nov 2025 (report date)
Transaction Fraud Report, parts 1, 2, 3 and 4 1 Jan 2026 (report date)
Account Balance Part 1 and Part 2 1 Jan 2026 (report date)
CSV transaction fees (related practice) 1 Jan 2026 (report date)
Shipping Cost Part 3 1 Dec 2025
Email Subscription Scheduler 1 Nov 2025 (report date)
Jupyter server logs (OA) 1 Nov 2025 (report date)

Project Coding

Question Reports Last seen
Complete a Bikemap Route and JSON Integration 2 Jan 2026
Mako debugging (practice) 1 Jan 2026 (report date)
JSONParser debugging 1 Jan 2026
requests task (onsite) 1 Nov 2025 (report date)

ML Modeling

Question Reports Last seen
Complete a Two-Label Tabular Model Pipeline 1 Nov 2025 (report date)

8 detailed candidate timelines, loops, and screens

Report 15 (reported February 2026, SWE New Grad: four CSV parts, three finished) is also covered in the Stripe new grad team-screen reports as entry S1.


16: Three shipping parts, then a rejected onsite

Report date: February 18, 2026 · Team screen: December 22, 2025 · Role: Full-time SWE New Grad · Result: Rejected January 27, 2026

  1. Team screen: Shipping Cost Part 1, Part 2 and Part 3, all finished. They fixed bugs along the way and wrote tests for parts two and three; part two took longest.
  2. January 20 onsite: the Java coding round was harder than expected and part two stayed unfinished. A JSONParser debugging round also stayed incomplete.
  3. Bikemap JSON Integration meant reading a JSON file, sending the request and checking the response.
  4. The rejection email gave no reason.

Report 17 (reported January 2026, SWE New Grad, India: a HackerRank screen with two engineers) is also covered in the Stripe new grad team-screen reports as entry 07.


18: Four fraud-report levels

Report date: January 15, 2026 · Role: Full-time SWE New Grad · Result: Passed this screen

  1. A 60-minute technical call with about 45 minutes of coding. Each level unlocked the next.
  2. Level 1: check that each six-field CSV row is complete.
  3. Level 2: amount-range and blocked-payment-method rules.
  4. Level 3: compare a transaction with the user's history and flag matches below 50%.
  5. Level 4: print at most two error codes in priority order, or OK, in an aligned report.
  6. The requirements were long to read but simple in logic. Not every original value or schema was shared.

19: An onsite map built from coordinates

Report date: January 15, 2026 · Level: Not confirmed · Stage: Virtual onsite

  1. Coding: Account Balance Part 1 and Part 2, covering transaction fees, nonzero balances and a list of rejections.
  2. Bikemap JSON Integration: a 400 × 600 map at zoom 13, centered near latitude 47.579, longitude -122.31. Points came from ride-simple.json. The first point was a start marker, the last a Stripe marker, and the route alternated ten blue points and ten orange points. The candidate combined this with earlier data, sent a POST request and saved the map.
  3. Mako debugging centered on a directory check and a node-visit method.

20: CSV transaction fees in a University SDE screen

Report date: January 8, 2026 · Role: University SDE (full-time new grad) · Status: Waitlist

  1. The candidate picked an open slot, then coded in a HackerRank sandbox with no required screen share; a local IDE was allowed with screen sharing.
  2. The task read CSV transactions and printed transaction ID, type, payment provider and fee. Sample output included py_1,payment,card,51, py_2,payment,card,0 and py_3,payment,klarna,101. The full fee table was not kept; the status and regional rate exercise is related practice.
  3. String handling and reading the rules took longer than the math.
  4. Afterwards the interviewer could ask about Stripe and take questions. No decision was given on the call.

Report 21 (reported May 2026, SWE New Grad, Bengaluru: a region-based tariff) is also covered in the Stripe new grad team-screen reports as entry 08.

Report 22 (reported February 2026, SWE New Grad, Dublin: strings, arrays and how to improve the code) is also covered in the Stripe new grad team-screen reports as entry 09.


23: Messages with malformed rows

Report date: January 2026 · Role: SWE New Grad

  1. The first interview felt like engineering work, not a LeetCode exercise. The program read from a file or standard input and processed messages.
  2. Edge cases included a row without the expected five tokens and an empty name. Knowing how to read and split input in their language saved time.
  3. The full rule, output, length and outcome were not shared.

24: A lowercase bug in col-verify

Report date: November 13, 2025 · Role: SDE New Grad · Length: About 45 minutes

  1. The candidate called the task col-verify, which is Project and Filter Structured Records: take a multi-line, multi-column string and return a filtered string as the conditions change.
  2. Part one went smoothly. In part two the tests caught a lowercase bug, which they fixed.
  3. They explained part three but did not finish coding it.

25: A Machine Learning Engineer loop

Report date: November 11, 2025 · Role: Machine Learning Engineer · Result: Did not pass

  1. Coding: Project and Filter Structured Records, selecting or changing column contents in a multi-line string. One bug cost a lot of time.
  2. A separate 45-minute round, Two-Label Tabular Model Pipeline: build a neural network in PyTorch or similar, handle null rows and missing features, and evaluate the model.
  3. Language docs were allowed; LLMs were not.

26: Shipping cost for the 2026 cohort

Report date: November 8, 2025 · Role: 2026 SDE New Grad · Result: Rejected

  1. The OA was about Jupyter server logs.
  2. The phone round was Shipping Cost Part 1 and Part 2; the candidate finished both and moved on.
  3. The onsite had Email Subscription Scheduler and a separate requests task.

A separate SDE report (level not given)

This report is outside the 8-report counts above because its level was not stated. One task moved from direct matches to weighted matches to a whole connected group.

S1: Linking related merchants

Interview date: September 23, 2026 · Report date: October 14, 2026 · Role: SDE · Result: Pending

  1. Each merchant had an ID, email, shop name and other attributes.
  2. Part one found merchants that share at least one attribute with the target.
  3. Part two gave each matching attribute a score and kept merchants at or above a threshold, as in Weighted Direct Links.
  4. Part three followed indirect links through other merchants, as in the full connected component. The exact input format, weights and examples were not shared.
  5. The candidate solved all three parts but lost time parsing input in C++, and said they would rather have used Python.

Patterns across the reports

  1. Records and CSV rows. Five of the 8 reports processed text or CSV records with growing rules (18, 20, 23, 24, 25).
  2. Shipping cost on the phone screen. Entries 16 and 26 both got Shipping Cost Parts 1 and 2 and advanced; 16 also finished Part 3. Both were later rejected.
  3. Reading beats logic. Entries 18 and 20 said the logic was the easy part; long requirements and string handling took the time.
  4. Bugs under the clock. Entry 24 fixed a lowercase bug the tests exposed, entry 25 lost time on one bug, and entry 16 fixed bugs while writing tests.

Common questions

How long is the Stripe new grad technical screen?

Entry 18 had a 60-minute call with about 45 minutes of coding, and entry 24 described about 45 minutes. The Stripe new grad team-screen reports have more timings, from 45 minutes to an hour.

What does the Stripe new grad phone screen ask?

Mostly record processing with layered rules: CSV checks, fees, shipping tiers or filtering columns (16, 18, 20, 24, 26).

Can I use my own IDE in the Stripe screen?

Entry 20 coded in a HackerRank sandbox and said a local IDE was allowed with screen sharing.

Does Stripe allow AI or LLMs in interviews?

In the MLE loop (25), language docs were allowed and LLMs were not.

Keep practicing

In this series: Part 1 · Part 2 (current)

Sources: Anywhere useful interview info can be found.

0 total reactions. Your reactions: none.
0 comments

Comments

0

No comments yet — be the first to share what you know.