EdgeCoach is a privacy-first iOS speaking coach that helps users practice presentations, interviews, and high-pressure speaking in real time. It tracks pace, filler words, pauses, tone, and optional eye contact, then turns each session into clear feedback and actionable next steps.
Built with SwiftUI and SwiftData, EdgeCoach extends that core experience with a real-time speech analysis pipeline focused on on-device performance and privacy. The app captures live audio, processes speech input, and evaluates delivery using metrics such as speaking pace, filler-word frequency, pause patterns, confidence trends, and vocal tone classification. It also supports optional camera-based eye-contact tracking, allowing sessions to combine voice and visual presence signals in a single coaching flow.
The training system is structured around multiple session types, including open practice, timed “Spotlight” runs, and scenario-based simulations for interviews, pitches, and Q&A. Each completed session is stored as a detailed metrics record and transformed into a post-session report with summarized scores, coaching insights, and targeted improvement goals. On supported devices, the app can enhance feedback further with on-device AI-generated content analysis using Apple’s Foundation Models.
Beyond session analysis, EdgeCoach includes a progression layer designed for retention and repeat practice: XP, streak tracking, achievements, mission systems, and daily reminders. From an engineering perspective, the app combines real-time media processing, lightweight ML classification, persistent session history, and adaptive feedback into a cohesive native iOS product aimed at deliberate communication training.