Say Again
Real-time clarification for a user with an auditory processing difference. Built for Say It Back — build a voice interface for someone who cannot comfortably use a screen. Live transcription with on-demand rephrase at three reading levels. The interface is deliberately narrow: one path, done properly, rather than a broad surface that half-works at the deadline.
Gallery
Screens
Say Again — main view
Say Again — result state
On the clock
What shipped, what didn't
Completed in time
Live transcription with on-demand rephrase at three reading levels.
Unfinished / faked
Struggles with more than two speakers.
Obstacles
Underestimated the parsing work. Cut a whole feature to protect the demo.
Lessons learned
Boring, well-understood tools beat interesting ones when there's no time to debug.
Process
Prompting approach
Started from a single system prompt describing the user and their constraint, then iterated with the assistant on the Say Again data model before writing any interface code. Full prompt log is in the repository under /prompts.
Fairness
Disclosures
Declared by the builder at submission time, visible to judges before scoring.
Started from the official framework starter template, unmodified.
Judging
Public feedback
Judges' private notes stay private — this is what they chose to say publicly.
Hana Kobayashi
Principal Research Engineer, Independent
Excellent use of the model — it is doing actual work here, not decorating a form. The failure case you showed earned you points rather than costing them.
Ravi Anand
Head of Design, Tessellate
Genuinely useful. I would install this. The limitations section is honest, which makes me trust the rest of the submission more.
Tomas Eriksen
Developer Advocate, Independent
Genuinely useful. I would install this. The limitations section is honest, which makes me trust the rest of the submission more.
Discussion
0 comments
No comments yet.