Practice: Cataloging Species in an Unfamiliar Environment
Solve a full decomposition practice problem in the style of Palantir's FDSE round, cataloging species in an unfamiliar environment. You'll scope a first version around a research lead's evening decision, model observations separately from identification decisions so original evidence survives relabeling, and keep incomplete records visible. The lesson also shows you how to design offline capture with idempotent uploads and versioned reviews, trace retries and revisions step by step, and define an honest pilot success measure.
We'll cover the following...
- What will your species catalog help someone decide?
- Which answers constrain your five-day version?
- Can your records retain O201, O202 and O203?
- What happens when O201 arrives twice?
- What survives O202’s changed identification?
- Comparison: What can the lead use each evening?
- Which consequential gap will you repair?
Prepare a first-version species-cataloging plan that helps a research lead review field evidence and choose tomorrow’s follow-up observations. Your plan must remain useful when labels change, phones are offline and evidence is incomplete. Write and briefly defend it in your own document before reading the comparison solution.
A candidate report describes a prompt to build something that gathers data about life on a planet; that supports this practice theme, not its exact wording, constraints or likelihood.
What will your species catalog help someone decide?
Start with an empty document. Before reading the answer packet, write:
Up to four clarification questions whose answers would change your design.
A provisional primary user and the decision your catalog supports.
Use a suggested 25-minute timebox for the whole attempt. This is a practice timebox, not a claim about interview format.
Pause here and write ...