Palantir · OA → Phone screen → Onsite
In processPalantir FDSE Intern OA and Interviews: 3 Experiences
Palantir FDSE intern OAs in 2023 and 2026: three linked discount problems with a rounding trap, or SQL, a paginated REST task and subarray products.
At a glance
- Process: a HackerRank OA, three tasks in 90 minutes (02); the 2026 OA took the candidate about 90 minutes (01). An older 2022 report (H03) adds the recruiter call, learning, decomposition and hiring-manager rounds.
- Asked recently: three building-block discount problems on one shopping cart (01); a SQL join and filter, a paginated REST average and counting subarrays with product at most
k(02). - Difficulty: both candidates found the OA manageable; 01 warned about rounding and 02 about the inclusive bound that differs from LeetCode 713.
- Practise first: Basic Item Discounts, Stacked Item and Category Discounts, Optimize Discount Selection and Freelance Candidate Review.
Coverage: 3 Palantir Forward Deployed Software Engineer (FDSE / FDE) intern candidates. Two are current: OA reports from November 2023 and September 2026. One 2022 New York commercial FDSE intern loop, ending in an offer, sits in History.
Other Palantir cohorts: New grad, Part 1 · New grad, Part 2 · Experienced hires
What came up
The two current OAs looked nothing alike: one cart-pricing problem in three steps (01), versus three separate SQL, REST and array tasks (02). Each topic below comes from one of the 2 reports.
| Topic | Reports | Last seen |
|---|---|---|
| Discount stacking, rounding and budget optimization | 1 of 2 | Sep 2026 |
| SQL joins, filters and sorting | 1 of 2 | Nov 2023 |
| Paginated REST API records | 1 of 2 | Nov 2023 |
| Counting subarrays by product | 1 of 2 | Nov 2023 |
Questions we found
All dates are report months.
Coding
| Question | Reports | Last seen |
|---|---|---|
| Basic Item Discounts | 1 | Sep 2026 |
| Stacked Item and Category Discounts | 1 | Sep 2026 |
| Optimize Discount Selection Under a Point Budget | 1 | Sep 2026 |
| Count subarrays with product ≤ k | 1 | Nov 2023 |
| Average pulse by diagnosis and doctor (paginated REST) | 1 | Nov 2023 |
SQL
| Question | Reports | Last seen |
|---|---|---|
| Freelance Candidate Review | 1 | Nov 2023 |
2 current candidate reports
01: Three discount problems with a rounding trap
Report date: September 15, 2026 · Role: FDE Intern · Stage: OA · Time: about 90 minutes (the candidate's estimate)
- All three parts share one shopping-cart setup. The candidate found it manageable; rounding was the thing to watch. The post header says internship, while a status field says recent graduate.
- Basic Item Discounts: items have names and integer-cent prices; coupons name an item and a percentage, and unmatched coupons are ignored. Return the subtotal, total discount and final total as integers. Round each item's final discount with Python's
round(), never the intermediate steps. - Stacked Item and Category Discounts: add category coupons such as 20% off Dairy. Apply the item coupon first, then the category one, multiplying what remains: 10% then 20% is 28% off. No item may save more than 80%; a coupon that would cross that is applied only up to 80%. At most one item coupon per item and one category coupon per category.
- Optimize Discount Selection Under a Point Budget: choose which coupons to turn on. Item coupons cost 2 points, category coupons 5, and the budget is 20. Maximize savings under the same stacking and cap rules, and return the chosen coupons and points used. The grader checks the output format before scoring, and this part allowed unlimited submissions.
02: SQL, a paginated medical API and an inclusive product bound
Report date: November 13, 2023 · Role: FDSE Intern · Platform: HackerRank · Time: 3 tasks in 90 minutes · Result: Passed
- SQL: Freelance Candidate Review. Join
profiles.idtostats.profile_id; returnfirst_name,last_name,emailandjob_success_scorefor verified profiles scoring at least 90, sorted by score descending, then first and last name, top ten rows. - REST API: write
pulseRate(diagnosisName, doctorId)over a paginated medical-record endpoint. Visit every page, keep records matching bothdiagnosis.nameanddoctor.id, averagevitals.pulseand truncate to an integer. Examples: 129 for pulmonary embolism with doctor 2, and 101 for common cold with doctor 2. The candidate found it easy in Python. - Subarray products: count non-empty contiguous subarrays whose product is at most
k. Up to 500,000 values from 1 to 100,kup to 1,000,000, answer as a long.[2, 3, 4]withk = 6gives 4;[1, 2, 3]withk = 4gives 4; withk = 7it gives 6. - The candidate used a linear two-pointer pass and stressed the inclusive "at most", unlike LeetCode 713's strict bound.
Common questions
What is on the Palantir FDSE intern OA?
In 2026, three linked discount problems on one cart (01). In 2023, a SQL query, a paginated REST API task and a subarray-product count, three tasks in 90 minutes (02).
What is the trap in the Palantir discount OA?
Rounding: round only each item's final discount, with Python's round(), and cap savings at 80% (01).
History (before Nov 2023)
This report is older than three years. It counts toward the total but stays out of the summary and tables above.
H03: An FDSE transfer, two technical rounds and an airline follow-up
Report dates: September 29 and October 25, 2022 · Role: FDSE Intern, commercial · Location: New York · Result: Offer
- The candidate first applied to SWE in Palo Alto (application August 9, OA August 19, Karat August 29); those questions stay with the SWE track. The recruiter call moved them to FDSE in New York.
- Recruiter (September 7): a self-introduction, why Palantir and SWE versus FDSE. A virtual onsite invite came September 14.
- Decomposition (one hour, September 27): Farm Pest Alert System. The interviewer gave pest and farm database schemas and asked how to build one feature (details withheld). Mostly spoken, with pseudocode or a schema in a text editor; solving the stated problem mattered more than naming database types.
- Learning (one hour): improve given code and API interfaces, mainly by making fewer API calls, for example by keeping results in a data structure. Then a new concept and a second snippet, compared with the first on time and space, explained in plain terms rather than Big O. Practice: Minimum Calls for Repeated API Lookups; the candidate did not mention LRU or a capacity.
- Both rounds began with 10–15 minutes of resume-based behavioral talk.
- Hiring manager (October 10, invite September 29): about 30 minutes behavioral, with deeper project questions on helping a peer, taking a risk and a tight deadline. Then about 30 minutes on an Airline Aircraft Assignment Planner: assign flights to aircraft at an airport; when an earlier flight is five hours late, send another aircraft and work out the new times.
- HR called October 11; the offer call and written offer came October 19.
Keep practicing
- Basic Item Discounts, Stacked Item and Category Discounts and Optimize Discount Selection: report 01's OA.
- Freelance Candidate Review: report 02's SQL task.
- Farm Pest Alert System, Minimum Calls for Repeated API Lookups and Airline Aircraft Assignment Planner: H03's onsite.
- Palantir coding bank, SQL bank and system design bank: all Palantir questions by type (sign-in needed).
- Palantir interview guide and Palantir hiring insights: stages, practice and recent hiring signals.
- Experience feed: interview reports from other companies and roles.
Sources: Anywhere useful interview info can be found.
Comments
0No comments yet — be the first to share what you know.