Amelia for the Deaf
The Persistent Context Sprint Hackathon | Live at MongoDB .Local Build FestAmelia The problem. Live captions get the words right, but words are the easy part. In a group conversation, deaf and hard-of-hearing people lose everything around them — who's speaking, which voice just changed their mind, what someone promised to do. Captions scroll away, and asking people to repeat themselves has a social cost hearing people never pay. What it does. Amelia listens from your phone and turns the conversation into something you can ask questions of, instead of a wall of text you have to keep up with. • Every line is tied to a real person, recognized by their voice. • She remembers each person, and tracks when things about them change. • She tracks promises in both directions — what you owe and what you're owed. • Say her name and she recognizes you by voice, answers from memory, and can speak the answer aloud to the room for you. Sponsor technologies. MongoDB Atlas stores everything and powers the search, using Atlas Vector Search for both voice recognition and memory. Anthropic Claude does the reasoning, Fireworks AI handles memory extraction, OpenAI transcribes and separates speakers, SpeechBrain makes the voiceprints, ElevenLabs is her voice, Resend drafts emails, and Mentra adds smart glasses. Built on Bun, Hono, and Expo.
