Second-Voice

Your voice. Your memories. Your choice. Second Voice is an AI-powered communication assistant designed for people living with ALS and others who have lost the ability to speak. It helps them participate in natural conversations without having to type every word. Traditional communication aids can speak what a user types, but they don't remember who they're speaking to, what they've discussed before, or how their communication changes between relationships. Second Voice does. Using MongoDB Atlas and Vector Search, Second Voice maintains relationship-specific memories, conversation history, personal phrases, and privacy preferences. When someone speaks, the system retrieves relevant context and generates personalized reply suggestions. The user chooses what to say with a single tap, and ElevenLabs turns that response into natural speech. What makes Second Voice different is memory the user can govern. If a suggestion feels too personal, the user can reject it. The system learns a privacy boundary and applies it to future conversations. For example, someone may comfortably discuss medication with their pharmacist while keeping the same information private from an acquaintance. Built with MongoDB Atlas, Nebius AI, Voyage AI, ElevenLabs, FastAPI, and React, Second Voice combines persistent memory, contextual intelligence, adaptive privacy, and accessible voice interaction. Our goal is simple: help people who have lost their speech retain their personality, relationships, privacy, and control over what they say.

Team