The Palantir Online Assessment (HackerRank OA)

The Palantir online assessment is a HackerRank test of roughly 60 to 90 minutes, and it is the interview's distinctive first gate. Its most-reported shape is three connected parts: one algorithm or implementation problem, one SQL query built on joins and aggregation, and one REST-API task where you page through an endpoint and parse the JSON. It is graded against hidden tests, so your code has to run and pass, and Python is the common pick. Some loops report a coding-only variant instead, so treat the three-part version as typical rather than guaranteed.

The OA is not a warm-up you can wing. It hits Forward Deployed candidates hardest, and the data-wrangling half is where many strong coders lose the gate. The coding overview covers the live rounds that follow; this lesson is the assessment itself, part by part.

The Three-Part Shape

The three parts often build on each other inside one sitting. The algorithm problem looks the most like a normal OA question, though it usually arrives wrapped in a short story you have to read the specification out of. The other two are the parts most assessments do not have, and they are the reason to prepare the Palantir OA specifically rather than treating it as generic HackerRank practice.

Part One: The Algorithm Problem

The algorithm problem sits at medium difficulty and almost always arrives as a scenario rather than a clean signature. The single most-repeated Palantir problem is a portfolio-holdings roll-up, where dated per-company share events aggregate into a total for each day. It pairs naturally with a non-zero-balance accounts problem that nets transactions and returns the accounts that do not settle to zero. Both are hash-map aggregation over a real-world story, which is the recurring core of the Palantir coding surface. Because it runs against hidden tests, handle the empty input and the boundary before you polish, and get both other parts to a passing state before you over-optimize any one.

The base techniques behind these problems are drilled in the free Foundations DS&A course; the Palantir-specific versions, with the follow-up twists stacked the way the assessment stacks them, are in Arrays, Strings & Hashing.

Part Two and Three: The Data-Wrangling Half

The Palantir angle

The SQL and API parts are where the Forward Deployed flavor lives, and they are the part most candidates neglect and lose the gate on. This is the day job of a deployed engineer wrangling a customer's real feed, so Palantir tests it directly rather than hand-waving it. The SQL problem is a genuine multi-table join with grouping and ordering, for example total session duration per city or a month-over-month change in ratings. If you have coasted on writing application code and let raw SQL go rusty, this is where it shows.

The API problem hands you a paginated endpoint. You loop every page, watch the total-pages boundary so you do not stop one page short or run one page long, parse the JSON, and aggregate the result. One reported version returns the university with the most international students across two cities, falling back to the second city when the first has none. Python keeps this part short, since JSON parsing and HTTP calls are a few lines, which is a large part of why it is the practical default for the whole assessment.

Common questions