Search⌘ K
AI Features

The Ideation Phase: Asking About Users, Data and Constraints

Run the ideation phase of Palantir's FDSE decomposition round by asking questions whose answers change your proposal. Identify who acts and what they can change, check data provenance, coverage, freshness and permissions, and keep uncertain identities and missing observations explicit. Let constraints such as delayed uploads and limited capacity reshape your scope, withdraw unsupported promises, and agree on a first version with a defined success measure and stakeholder guardrails.

“Help us reduce flooding”: Which question earns the next minute, “Which cloud should we use?” or “Who acts on this information, and what can they change?” The second can determine the entire solution boundary. Clarification earns its time when the answer changes a decision you would otherwise make. Your aim is to reach a useful first version whose promises fit the evidence and operational limits.

Which question changes the flooding proposal?

Choose the opening question before seeing who can act.

Which Question Earns the Minute?

Q1. Your only brief is “Help us reduce flooding.” Which opening question earns the next minute?

Do you find this helpful?

The coordinator’s limited authority also changes what “useful” means: information must help choose tomorrow’s visits, not merely describe flooding. A technically accurate citywide display could still fail to support that decision.

Use this private test: Name two plausible answers and say what you would do differently. A user who dispatches emergency crews needs a different workflow from someone planning tomorrow’s inspections. Cloud preference can wait until it affects a real constraint. Permissions and safety questions can also establish mandatory boundaries, even when they do not change which locations rank highest. ...