I Built a Private AI Pronunciation Coach So My Friend Wouldn't Give Up on Her Placement Dreams
Prize Categories Overall Winner Best Use of Gemma 💡The Raw Truth: Why I Built This We are currently sitting for our final-year campus placement drives. The pressure in the air is suffocating, but for my closest friend, it has been downright terrifying. She comes from a rural vernacular background. In our computer science labs, she is a rockstar—her code is elegant, her logic is sharp, and her…
Prize Categories Overall Winner: "Best Use of Gemma"
The author, who hails from a rural background, built a private AI pronunciation coach to support her friend during campus placement drives. Her friend, despite excelling in coding and problem-solving, struggles with speaking English due to her accent and regional vowel shifts.
During an automated AI assessment, her voice was flagged as "unclear" due to her accent, leading her to doubt her abilities and contemplate leaving corporate jobs. The author promised her friend a better solution, one that could provide empathetic guidance without leaving her laptop.
The "Placement Pronunciation Coach" is an acoustic-backed, locally-run AI speech evaluation dashboard designed specifically for students facing interview anxiety. It uses a hybrid acoustic pipeline that includes:
1. Target Prompt Alignment: Compares expected interview responses against recorded user audio.
2. Acoustic Ambiguity Detection: Uses OpenAI Whisper with word_timestamps=True to identify words where speech clarity drops (threshold at 0.75).
3. Phonetic Sound Alignment: Converts expected and spoken text into IPA/Arpabet phonemes to detect real acoustic mismatches.
4. Empathetic Coaching via Gemma 2:2b: Empathetic system prompts from Google Gemma 2 generate supportive communication scorecards, natural interview rewrites, and simple syllable sound-outs.
The tool runs entirely locally, ensuring privacy, offline accessibility, and customized empathy for the user.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
