whisper-large-v3

Local-first macOS app that records lectures and meetings, transcribes audio on-device with Whisper large-v3 (Apple Silicon MPS), and generates structured Obsidian markdown notes via Gemini AI, your audio never leaves your Mac

Voice-first AI German fluency trainer - multi-agent (LangGraph), RAG (LlamaIndex+Qdrant), FSRS learner model

A Kaggle-native, multi-GPU asynchronous AI framework for Egyptian Arabic psychiatric speech analytics. Integrates late decision-level multimodal fusion (Wav2Vec2 + Whisper V3 Large + CAMeLBERT) with Qwen-powered clinical report generation.

A Kaggle-native, multi-GPU asynchronous AI framework for Egyptian Arabic psychiatric speech analytics. Integrates late decision-level multimodal fusion (Wav2Vec2 + Whisper V3 Large + CAMeLBERT) with Qwen-powered clinical report generation.

Desktop app to transcribe YouTube videos using yt-dlp, ffmpeg, and Groq Whisper. No server, no terminal.