Search⌘ K
AI Features

The Implementation Coding Problem

Tackle the implementation coding problem in the Palantir FDSE screen by turning an ambiguous, practical prompt into a precise contract and a small working baseline. Clarify ordering, state and boundary rules, work out expected results before running any code, and write a readable Python function supported by a loop invariant. Design discriminating tests that expose plausible bugs, justify time and space complexity, and report only the checks you actually ran.

Two requests for three pens each arrive when only five pens are available. “Fulfill these requests” could mean giving the second requester two pens, checking both requests against the original five, or fulfilling only the first request. Resolving that ambiguity makes a small implementation possible within a screen timebox. A working baseline is the simplest implementation that satisfies the whole agreed contract.

What does “fulfill these inventory requests” mean?

A function contract states accepted inputs, required outputs, ordering, state changes and behavior at relevant boundaries. Ask: “Are requests processed sequentially? Can fulfillment be partial? Should this function change the supplied stock?”

Use these agreed rules:

  • Process requests in supplied order, using the stock remaining after earlier fulfillments.

  • Fulfill the whole requested quantity when current stock is sufficient, including equality. Otherwise, leave stock unchanged.

  • Return one (request_id, status) tuple per request, in the same order. Status is fulfilled, insufficient_stock or unknown_sku.

  • Keep the caller’s stock mapping unchanged.

Input validity is guaranteed: Stock quantities are nonnegative integers; requested quantities are positive integers; identifiers are nonempty strings; request IDs are unique; and inputs have the promised shape. These guarantees remove the need to invent a validation subsystem. Zero stock still needs defined behavior because it is a valid input.

Palantir’s coding guidance, as reproduced in a secondary compilation, emphasizes turning an algorithm into working code, considering edge cases, and explaining time and space complexity. One secondary preparation guide recommends clarifying underdefined practical prompts and getting a clean solution working before refining efficiency; this is preparation advice rather than a guaranteed assessment format or official scoring rubric.

Which expectations can I write before coding?

A fixture is the concrete input data used in a check. An executable specification checks actual behavior against an independently determined expected result.

Start with stock {‘pen’: 5, ‘book’: 2} and these requests, in order:

('R1', 'pen', 3)
('R2', 'pen', 3)
('R3', 'pen', 2)
('R4', 'map', 1)
('R5', 'book', 2)
('R6', 'pen', 1)
Six ordered requests used as fixture data

The repeated pen requests distinguish sequential behavior from independent checks against original stock. Predict what remains afUter R2 and whether R3 succeeds. ...