What Muse does not know yet.

• What should be the first genuinely valuable Muse experience?
• How much of wardrobe management should be automated?
• How much control should remain with the user?
• What is the smallest useful version of Muse?

• What is the lowest-friction way to digitize a wardrobe?
• How should users correct incorrect AI interpretation?
• How should recommendations be presented?
• When should Muse explain why it made a recommendation?

• Which visual attributes can be reliably extracted?
• How much can general-purpose AI handle?
• Where are specialized models necessary?
• How should uncertainty be represented?

• How robust is clothing recognition across different photography conditions?
• Is exact product identification actually necessary?
• How useful is visual similarity on its own?

• What is the minimum data model for a useful garment?
• Which attributes should be automatically generated?
• Which should be user-defined?
• How should personal preferences be represented?

• What is the most valuable product experience?
• Who is the initial paying customer?
• Which parts of Muse could support a sustainable business?

• How does Muse differentiate from existing fashion and wardrobe products?
• Is the strongest value in wardrobe organization, outfit creation, recommendation, or something else?

• How can Muse feel intelligent without becoming complicated?
• How should fashion aesthetics coexist with technical credibility?

A question leaves this page only when it is genuinely resolved — and when it does, it is linked to the decision or experiment that resolved it.