Daniel Rodriguez

Context

Meta Display Glasses introduced a new kind of private text input. With the EMG bracelet, users could write through natural finger movements on a surface and see text appear through the glasses display. The ask came in late and carried real organizational pressure, including direct executive attention. The team needed to move in under two weeks from unclear interaction options to prototypes that could support research, engineering, and a default recommendation.

Challenge

The system needed to make users feel confident while writing, but not so focused on every character that they stopped trusting correction behavior. Per-letter feedback gave users the most visibility, but it encouraged them to watch every character and intervene too early. Hidden or delayed feedback protected flow and privacy, but it asked users to trust a new input method with very little confirmation.

What I shaped

I built prototypes for per-letter, per-word, and more hidden feedback models, then helped evaluate how each affected confidence, privacy, correction behavior, and flow.

• Per-letter feedback was explicit, but made users hang on every character.
• Hidden output protected attention, but felt too opaque as a first-use default.
• Per-word feedback gave enough signal while encouraging users to keep writing.

Outcome

The prototypes helped unblock research and engineering decisions for EMG handwriting on Meta Display Glasses inside an under-two-week push. Per-word feedback became the shipped default because it gave users a useful level of confirmation without making them over-monitor every character.

Reflection

New hardware inputs are not solved by making every state more visible. Sometimes the better default is the one that gives people enough confidence to keep moving.