Start from a real problem that real people have

The strongest proposals begin with a sentence like "Small pharmacies in our area track expiry dates on paper and throw away stock every month", not "We want to build an app using Flutter and Firebase". A real problem gives you users to talk to, a way to measure success, and a story examiners remember.

Good places to find one:

• People you know: a family business, a clinic, a school, your university's own offices.
• Your own frustrations: something slow, manual or error-prone you have dealt with yourself.
• Open datasets and public services: problems where data exists but nobody has made it useful.
• Faculty research: a supervisor's ongoing work often has a well-defined piece a student team can own.

Write the problem down in one or two plain sentences before you discuss solutions. If you cannot, you do not have a problem yet, only a technology you want to use.