Design sprints
Work through an unclear product decision before your team commits to building it.
Start a projectMap user flows
QueuedDefine success metrics
QueuedNatural-language finder
ActiveEmpty + error states
ActiveResearch synthesis
DoneStakeholder kickoff
Done- Framed problem statement + success metrics
- Concept sketches and option exploration
- Just-enough screens to test the direction
- Feedback from available users or stakeholders
- A clearer direction ready for further testing
- Decision log so the "why" survives
- 1
Understand
Map the problem, constraints and what success means.
- 2
Diverge
Generate competing directions instead of one safe bet.
- 3
Decide + test
Commit to a direction and mock it up just enough to test.
- 4
Validate
Test with available users or stakeholders when possible.
Desearch: the open question was how to make AI answers trustworthy. Instead of adding disclaimers, the chosen direction put the sources beside every answer so people could check them.
A short, focused effort that turns uncertain requirements, competing priorities, and early ideas into a clear product direction you can test.
One to two weeks. If the direction is still unclear at the start of a larger project, we can use the first week or two this way before moving into interface design.
A framed problem statement with success metrics, concept sketches and option exploration, just-enough screens to test the direction, feedback from available users or stakeholders, and a decision log so the reasoning survives.