-
-
Switching ai models
-
Connect with external tools
-
Changin ai persona
-
Main Menu
We started this project after noticing a failure pattern in school and work: the most important decisions happen in fast conversations, and then vanish because nobody has time to write them down in the moment. We learned that the real problem is not access to information, but the cost of capturing context while life is happening. So we built a real-time artificial intelligence voice assistant with a one-second phone trigger that captures the right slice of audio and outputs structured outcomes: key points, decisions, responsibilities, and next actions. We added persona masks so the same facts can be prioritized for different roles, like student, teammate, or manager. We built the first prototype by wiring audio capture to a lightweight processing pipeline, then testing daily with classmates and coworkers. The hardest challenges were latency, noisy environments, and trust: people need speed, accuracy, and privacy boundaries, or they will not use it.
Log in or sign up for Devpost to join the conversation.