🗓️
iOS
PTer
This app provides a comprehensive report to support effective presentation practice. It analyzes whether your voice volume is appropriate, your pronunciation is clear, and your speaking speed is too fast or too slow. By repeating the same presentation multiple times, users can also generate comparison reports to track how much they’ve improved.
My Role
Design & Planning
Wireframing
UX flow design
Research on sound-related technologies
Research on Apple technologies
Development
Speech-to-text (STT) using Apple’s Speech Framework
Implemented Korean STT
Pronunciation clarity evaluation based on model confidence
Overall view implementation
Project folder structure setup
Backlog creation
Tech Stack
Figma
Xcode + SwiftUI
AVFoundation
Speech (Apple Speech Framework)
WhisperKit
Lottie
GitHub
Technical Overview
This project was built with the goal of leveraging Apple technologies to their fullest, with Sound as the core theme.
We primarily used AVFoundation for audio/video playback and editing, and incorporated Apple Speech and WhisperKit for speech-to-text functionality.
Challenges & Solutions
One major challenge was figuring out how to evaluate pronunciation clarity.
WhisperKit provided accurate results but was too slow, and the ETRI (Electronics and Telecommunications Research Institute) API allowed only one sentence per request, limiting usability.
In the end, we chose to use Apple’s Speech Framework, evaluating pronunciation clarity based on the confidence values returned by the model.