Skip to content

How it works

Stable language. Adaptive context. Never both in the same place.

EyeSay splits a communication system into the part that must never change and the part that should. That separation is the whole design.

1. The Voice Canvas never moves

Forty-eight permanent slots in an 8×6 grid. Each slot has a stable identity, so a word learned in row 2, column 5 stays in row 2, column 5. Category cells that open more words are shaped and marked differently from cells that speak, so the difference is never carried by colour alone.

2. The Context Halo carries the change

Eight visually distinct slots sit alongside the grid. They respond to time of day, the current routine, the place and the communication partner. The halo is frozen while a selection is in progress, so nothing shifts under a finger or a dwell. It always includes SOMETHING ELSE, so a suggestion set is never a closed choice.

3. The Agency Strip is permanent

STOP, NO, HELP, BREAK and TOILET are always reachable and cannot be removed by routines, editing or context. Self-advocacy is not a reward for compliance.

4. Nothing speaks without a press

Selected words gather in the utterance ribbon with Undo, Clear and a large SPEAK. Longer sentence options can be offered, but a suggestion only becomes speech when the communicator presses it — and SPEAK MY WORDS is always offered alongside, unchanged.

5. Partners model, they do not speak for

Model Mode mirrors the same grid topology and highlights cells for teaching. It cannot insert anything into the communicator's ribbon and it cannot trigger their voice.

What EyeSay does not do

  • It does not infer emotion, behaviour, competence or hidden intent.
  • It does not send the communicator's messages to a cloud service.
  • It does not diagnose, treat or cure any condition, and makes no clinical claims.
  • It does not reorder learned vocabulary to improve a prediction score.
Open EyeSay