Presentations rarely fail because people don’t know the content. They fail when pressure disrupts pacing, attention, and confidence.
This project trains regulation under evaluation by repeatedly rehearsing anxiety-inducing cues in a controlled environment.
- Trains: staying coherent, engaged, and recoverable under pressure
- Does not train: content quality, charisma, or accent/perfect delivery
The hardest part was not adding realism. It was deciding how computation should be structured so the experience supports real training.
I designed a repeatable loop where users can re-enter pressure, observe responses, and target one concrete improvement each run.
KEY OUTPUTS
A small set of regulation-focused signals + next-run goal.
Event scheduling, signal logging, and feedback UI in Unity.