2026 Rippling AI Coding Interview, Experienced SWE Phone Screen / Onsite: 9 Experiences
- Company
- Rippling
- Reports
- 9 interview reports
- Level
- Full-time
- Stages
- Phone screen → Onsite
Rippling runs its AI coding interview inside HackerRank with the platform's own assistant. The phone-screen tasks were Rippling problems with a long history, and the onsite AI round kept coming back to one HR codebase.
Coverage: 7 experienced full-time software engineer reports (one Senior Software Engineer title, one product-side role, the rest general SWE), plus 2 full-time reports that don't state a hiring level, kept in a separate supplement · Technical phone screens and onsite AI coding rounds, with the other onsite rounds where candidates described them · Interviews July–December 2026, reports posted October 27 – November 6, 2026 · Locations not stated
What came up
Five of the seven experienced reports place Rippling's AI coding round in HackerRank's AI-enabled environment. Candidates wrote less about the problems themselves and more about which AI they were allowed to use, how they steered it, and how they explained what it wrote.
Every experienced report named a task, and each one maps to a Rippling problem FastPrep already tracks.
| Topic | Reports | Most recent report date |
|---|---|---|
| HackerRank AI-enabled environment | 5 of 7 | Nov 6, 2026 |
| Whether AI was optional, or which AI was allowed | 3 of 7 | Nov 6, 2026 |
| Logger with message handlers | 2 of 7 | Nov 6, 2026 |
| Unit tests written during the round | 3 of 7 | Nov 4, 2026 |
| Camel Cards hand comparison | 2 of 7 | Nov 4, 2026 |
| Onsite HR codebase: scheduled employee changes | 2 of 7 | Nov 4, 2026 |
| Delivery cost tracker with rate changes | 2 of 7 | Nov 3, 2026 |
| Steering the AI with code comments | 2 of 7 | Nov 3, 2026 |
| Expense rules engine | 1 of 7 | Oct 29, 2026 |
These counts cover the 7 experienced reports only. One candidate can appear in several rows.
Questions we found
Counts below are for the 7 experienced reports. Dates from the two supplement reports and from other Rippling sightings already on FastPrep are labeled separately and don't add to the counts.
AI Coding: phone-screen problems
| Question | Frequency | All last seen dates |
|---|---|---|
| Stateful Log Transform and Search (logger with handlers, then search) | 2 reports | Reports: Nov 6, 2026; Oct 27, 2026 (report dates). Additional FastPrep history: Oct 19, Oct 7, Aug 30, Jul 29, 2026 |
| Camel Cards (which of two hands wins) | 2 reports | Reports: Nov 4, 2026; Nov 3, 2026 (report dates). Additional FastPrep history: Aug 2, May 29, Mar 3, 2026; Dec 22, 2025 |
| Delivery Cost Tracker with Effective-Dated Rates | 2 reports | Reports: Nov 3, 2026; Nov 2, 2026 (report dates). Additional FastPrep history: Oct 26, Oct 23, Oct 12, Sep 6, Aug 21, 2026 |
| Best and Worst Camel Card Completions (Camel Cards part 2) | 1 report | Reports: Nov 3, 2026 (report date). Additional FastPrep history: Oct 6, Sep 29, 2026 |
| Corporate Card Expense Rules (expense rules engine) | 1 report | Reports: Oct 29, 2026 (report date). Additional FastPrep history: Oct 13, Sep 18, Sep 1, Aug 5, Jun 29, May 16, 2026; Dec 31, Sep 16, Jun 16, Jun 3, 2025 |
AI Coding: onsite project on an existing codebase
| Question | Frequency | All last seen dates |
|---|---|---|
| Schedule Future Employee Changes in an HR API | 2 reports | Reports: Nov 4, 2026; Oct 29, 2026 (report dates). Supplement reports: Nov 5, 2026; Oct 30, 2026 |
System Design
| Question | Frequency | All last seen dates |
|---|---|---|
| Design a Product Analytics System (user-behavior and metrics aggregation) | 2 reports | Reports: Nov 4, 2026; Oct 29, 2026 (report dates). Supplement report: Oct 30, 2026 |
Behavioral (BQ)
| Question | Frequency | All last seen dates |
|---|---|---|
| Tell me about a time you gave critical feedback to someone | 1 report | Oct 29, 2026 (report date) |
7 detailed candidate timelines, loops, and screens
01: The logger question, and an assistant that only helped with syntax
Report date: Nov 6, 2026 · Interview period: October–December 2026 · Role: Experienced full-time SWE · Stage: Technical phone screen · Platform: HackerRank · Result: Round stopped partway; the candidate was rescheduling
Technical phone screen
- The task was the multi-part logger problem the candidate called the standard one, matching Stateful Log Transform and Search.
- The candidate opened with edge-case questions they had prepared. The interviewer seemed to think these were unnecessary and simply listed the common cases himself.
- Once coding began, HackerRank's built-in AI would only offer syntax help. The interviewer told the candidate to use their own AI instead, but they had nothing set up on their computer. The round was cancelled partway through.
- While rescheduling, the recruiter confirmed the opposite: if AI is used, it has to be HackerRank's AI.
The candidate's worry afterward was simple. If the built-in assistant really does that little, writing every part of the logger by hand in one round looked unrealistic.
02: Camel Cards on the phone, HR scheduling at the onsite
Report date: Nov 4, 2026 · Interview period: October–December 2026 · Role: Experienced full-time SWE, product-side position · Stages: Phone screen, onsite · Result: Rejected the day after the onsite
Phone screen
- The problem was Camel Cards, a poker-style hand comparison.
- AI use was optional here, and the candidate chose to use it. They explained their plan, walked the interviewer through what the AI produced, and also had it write a large batch of unit tests.
- The interviewer was satisfied, and the candidate moved on to the onsite.
Onsite: AI coding
- The candidate got an existing codebase for an HR system where staff edit an employee's salary, level, title and similar fields. The job was to add a scheduled-update feature, the same task as Schedule Future Employee Changes in an HR API. They believed it was a new question.
- The round used HackerRank's built-in assistant. GPT 5.5 was among the models offered, and the candidate felt that was good enough for an interview.
Onsite: system design
- A user-behavior and metrics monitoring aggregator, close to Design a Product Analytics System. The candidate saw it as an infrastructure-leaning question for a product role, hadn't prepared it well, and thinks they failed it.
Onsite: hiring manager
- The candidate presented a slide deck about a past project, followed by a few basic behavioral questions. The manager seemed only half-interested, and the candidate expected a fail here too.
03: A delivery-fee design solved in one shot, then questioned
Report date: Nov 3, 2026 · Interview period: October–December 2026 · Role: Experienced full-time SWE · Stage: Technical phone screen · Platform: HackerRank · Result: Rejected
Technical phone screen
- The task was the delivery-fee system design the candidate recognized as a known Rippling problem, matching Delivery Cost Tracker with Effective-Dated Rates. The prompt was long and already covered most edge cases; reading it took five to ten minutes.
- The strongest model in HackerRank's built-in AI appeared to be "GPT 5.6 terra", as the candidate read it.
- The interviewer first asked for the overall structure and data structures, then for the kind of timestamp to use and how a rate update would work.
- The candidate explained the plan while writing it out as comments, aiming for the AI to produce the whole solution in one pass. It did, and the interviewer then asked why the code was built that way.
The feedback was that they should arrive at the solution step by step and engage with the interviewer more along the way.
04: Camel Cards in two parts, with no say on the AI
Report date: Nov 3, 2026 · Interview date: a Friday shortly before the post (October–December 2026) · Role: Experienced full-time SWE · Stage: Technical phone screen · Platform: HackerRank · Result: Passed; onsite invite the following Monday
Technical phone screen
- HR had said AI would be optional in this round. The interviewer opened by calling it an AI coding round, and no choice came up. The candidate had planned to use AI anyway.
- Part 1 asked which of two hands is stronger, as in Camel Cards.
- Part 2 added the best and worst hands, the follow-up in Best and Worst Camel Card Completions. The AI wrote a brute-force version, and the two of them talked through how to make it faster without coding it.
- The candidate wrote their own reasoning as code comments and had the AI implement from those. They found the HackerRank assistant capable and thought it was worth getting used to its interface beforehand.
05: Senior SWE, a food-delivery simulation from a blank file
Report date: Nov 2, 2026 · Interview period: October–December 2026 · Role: Senior Software Engineer · Stage: Technical phone screen, rated hardest by the candidate · Platform: HackerRank, AI-enabled · Result: Rejected
Technical phone screen
There was no starter code. The candidate defined the classes and functions from scratch, in Python, for a system that tracks delivery drivers' pay rates and deliveries and computes the cost, close to Delivery Cost Tracker with Effective-Dated Rates. The requirements and follow-ups came in this order:
- Store driver rates and delivery records in dictionaries, using Unix timestamps.
- Support rate updates that take effect at a given timestamp and change how deliveries are costed.
- Write their own test cases and run them.
- Look up and use a driver's historical rates.
- Handle money precision: explain why using floats and rounding to two decimals isn't enough, and how to do it properly.
- Bring
getTotalCost()down from O(n) to O(1).
The report doesn't say how much of the code the AI wrote.
06: Expense rules on the phone, a scheduling argument at the onsite
Report date: Oct 29, 2026 · Interview period: October–December 2026 · Role: Experienced full-time SWE · Applied: Online · Stages: One phone screen, three onsite rounds · Result: Rejected
Phone screen: AI coding
The problem was the expense rules engine the candidate knew from earlier reports, as in Corporate Card Expense Rules. The follow-ups were less familiar:
- With a million expenses and 10,000 rules, how would you speed it up? The candidate said batch and aggregate, and treated basic rules and group rules separately.
- How would you store the results and notify customers?
- How would you optimize compound rules when one rule contains another, such as rule A-and-B where B is itself A-and-C?
The onsite invite came three business days later.
Onsite round 1: system design
- Design metrics aggregation, the event pipeline in Design a Product Analytics System. There were many events across many types, on the order of a million events and 10,000 event types.
- Requirements: collect the events, power dashboard queries like clicks in a day, add context to events (for example location derived from the client IP), and support heavier queries like clicks in the US over the last three months. Time ran out before the fourth requirement.
Onsite round 2: AI coding (Web API)
- An HR system whose database holds employee data and employee history. The new feature was scheduling changes ahead of time, such as a promotion with a title change next month. This is Schedule Future Employee Changes in an HR API. Frontend and backend both change, with the focus on the backend.
- The candidate first proposed a new table plus a backend worker that picks up due changes and applies them. The interviewer didn't accept that.
- They switched to storing the changes in history and applying the ones that are due when data is read, making sure a newer change doesn't overwrite fields from an older one.
- The design discussion took so long that there was no time left to fix bugs. The candidate thinks this round is where they failed.
Onsite round 3: project deep dive
- A three-slide deck about a past project, with many trade-off questions.
- One behavioral question: a time they gave critical feedback to someone (practice).
07: Logger handlers, then search
Report date: Oct 27, 2026 · Interview period: July–September 2026 · Role: Experienced full-time SWE · Applied: Online · Stage: Technical phone screen · Result: Listed as "Other"
Technical phone screen: AI coding
- The candidate first agreed the exact behavior of each handler with the interviewer, then asked the AI to implement it, then asked the AI for unit tests.
- The task, matching Stateful Log Transform and Search: a logger sends every message to a set of handlers. The problem itself defined four handler types: one removes every occurrence of a configured string and prints the result, one keeps only the first N characters and prints, one uppercases and prints, and one stores the message in an internal list without printing.
- Follow-up: search the stored messages.
- Discussion follow-up: how to handle messages that contain punctuation.
Supplement: hiring level not reported
These two full-time reports don't state whether the candidate was experienced, new grad or something else, so they sit outside the counts above. Both are onsites, and both got the same HR scheduling task.
08: An onsite where effective dates ate the review time
Report date: Nov 5, 2026 · Interview period: October–December 2026 · Role: Full-time SWE, level not stated · Stage: Onsite · Result: Rejected
AI coding
- An HR app with a frontend, backend and database, where editing an employee (for example setting a salary) takes effect immediately. The task was to let an edit carry a future effective date and only apply on that date, as in Schedule Future Employee Changes in an HR API.
- The interviewer kept pressing on how effective dates should be handled. The candidate compared resolving changes on read versus on write.
- Because the conversation ran on, they had little time to review the code the AI wrote, and they suspect this round cost them the offer. Report 06 describes a similar squeeze.
System design
- A news aggregator, in the same family as Design a Personalized News Aggregation Platform. The candidate found the interviewer easy to talk to.
Project deep dive
- A full hour with a director, with plenty of questions along the way.
The candidate mentioned writing up their phone screen separately; that report isn't identified here.
09: The most detailed spec of the HR scheduling task
Report date: Oct 30, 2026 · Interview period: October–December 2026 · Role: Full-time SWE, level not stated · Stage: Onsite · Result: Rejected
Round 1: AI coding (Web API)
The candidate found this round manageable and the interviewer very friendly. The task was Schedule Future Employee Changes in an HR API, with these rules:
- Each employee has an Edit tab. Change it so a user can schedule a future change to manager, department, salary or title.
- After submission, the scheduled change shows up in the employee's History tab.
- Use AI to change both the frontend and the backend.
- Don't hold future changes in memory. Reuse the existing EmployeeHistory table.
- Validate the effective date: not before the hire date and not after the termination date.
- Cancelling a scheduled change was out of scope.
- Two changes effective on the same day: the later one replaces the earlier one only for the same field. A salary change and a title change on the same day merge.
Follow-ups: supporting backdated changes, guaranteeing a scheduled change really applies on its date, and production concerns such as concurrency, idempotency and error handling.
Round 2: project deep dive
- The candidate brought a few slides, but the interviewer seemed uninterested from the start and asked very little. The project part ended after about half an hour, a few behavioral questions followed, and the hour-long slot finished before 50 minutes. The candidate blames this round for the rejection.
Round 3: system design
- A product analytics system for a startup that wants to track user behavior without a third-party tool, the question in Design a Product Analytics System.
- Pieces: an SDK for mobile and web apps and an API to receive their events; dashboards showing event counts per minute, with queries reaching back a week answered within 500 ms at the 99th percentile; event enrichment such as reverse geo lookup; and a warehouse for heavier queries like page loads by US state over three months.
- Scale: about 1 billion events a day, no traffic spikes to plan for.
- Follow-ups: events arriving out of order, deduplication, and keeping latency low.
Patterns across the reports
These cover the 7 experienced reports.
-
HackerRank's own assistant, of uneven strength. Reports 01–05 place their AI round on HackerRank, and 01–04 name its built-in AI. Report 02 saw GPT 5.5 among the models and report 03 saw "5.6 terra". Reports 02 and 04 found it good enough; report 01 got one that only helped with syntax.
-
Mixed messages about the AI. In 02, AI was optional and the candidate opted in. In 04, HR called it optional but the interviewer ran it as an AI round. In 01, the interviewer said to use their own AI, and the recruiter later said only HackerRank's AI was allowed.
-
Phone screens reused Rippling problems with a long history. Every phone-screen task matches a Rippling problem FastPrep had seen before these reports, and reports 01, 03, 04 and 06 say they already knew theirs.
-
Comments first, then the AI. Reports 03 and 04 wrote their plan as code comments and let the AI implement it. Report 04 passed. Report 03 got the whole solution in one pass and was told to build it up step by step and talk with the interviewer more.
Keep practicing
- Stateful Log Transform and Search: the logger from reports 01 and 07.
- Camel Cards and Best and Worst Camel Card Completions: both parts from reports 02 and 04.
- Delivery Cost Tracker with Effective-Dated Rates: reports 03 and 05.
- Corporate Card Expense Rules: report 06.
- Schedule Future Employee Changes in an HR API: the onsite HR project from reports 02, 06, 08 and 09.
- Design a Product Analytics System and Design a Personalized News Aggregation Platform: onsite system design from reports 02, 06, 08 and 09.
- Rippling question banks (sign-in needed): project coding, coding, system design and behavioral.
- Rippling Senior SWE Phone Screens / Onsites: earlier senior loops. Rippling SWE Intern OA / Interviews: the intern cohort.
- Rippling interview guide, Rippling hiring insights and the interview experience feed.
AI coding rounds at other companies
- Meta: Meta AI coding interview: two onsite rounds where you finish a hard algorithm inside a mostly written codebase with an AI model, then face stress-test suites.
- Google: Google AI coding interview: an online assessment with HackerRank's built-in assistant, and a live round on a multi-file project where using AI was optional.
- Stripe: Stripe AI coding interview: experienced engineers kept getting the same transaction rule parser, built step by step with a built-in model.
- Stripe interns: Stripe intern AI coding round: ten intern accounts of the same AI-assisted round.
- DoorDash: DoorDash AI coding interview: AI CodeCraft: build and run several small services live with your own AI tool.
- LinkedIn: LinkedIn AI coding interview: one long, multi-part task in an IDE whose agent edits files directly.
- Shopify: Shopify AI coding interview: pair programming with your own AI on Shopify's well-known question set.
- Datadog: Datadog AI coding interview: bring any AI tool and share your screen; the tasks are easy, and the follow-ups decide the round.
- Capital One: Capital One AI coding interview: new grad Power Day agentic round: fix a bug, then add a feature to a large full-stack app with Claude Code.
- Bloomberg: Bloomberg AI coding interview: intern rounds that start in a small file and move to a large codebase, debugging with AI help.
- xAI: xAI AI coding interview: new grad take-homes and onsites where you build with AI.
- Adobe: Adobe AI coding interview: AI showed up as one step in a director round and in a take-home; the rest stayed classic.
- Snowflake: Snowflake AI coding interview: design and build a rate limiter or a retrying task queue while the AI writes the code; one AI Engineer onsite was in person with your own laptop.
- Snap: Snap AI coding interview: a backend-only chat service from an empty Python project with Claude or Codex, or a feature in an existing codebase.
- Ramp: Ramp AI coding interview: build inside a repo Ramp provides; one repo logged every prompt for the interviewer to read.
- Anthropic: Anthropic AI coding interview: review four pull requests, then improve one with Claude Code.
- Apple: Apple AI coding interview: CoderPad with a sidebar agent, alongside Apple's other recent interviews.
More companies with AI rounds on FastPrep:
- Amazon: Amazon interview report: an AI-assisted debugging OA (1 experience).
- Atlassian: Atlassian interview report: project coding and AI tool use (11 experiences, two parts).
- Circle: Circle interview report: three AI rounds: building, debugging and PR review (1 experience).
- Volta: Volta interview report: a deliberately vague data task with a token budget (1 experience).
Across companies: 2026 AI Coding OAs at 15 Companies: 32 Reported Problems and 2025–2026 AI Coding Phone Screens and Onsites at 43 Companies: 76 Reported Problems.
Sources: Anywhere useful interview info can be found.
Comments
0No comments yet — be the first to share what you know.