The problem
An AI agent in a group conversation hears one stream of words. Without knowing who said what, it can’t track who asked, who agreed, or who it should answer.
What I built
Real-time speaker identification that tags each voice in a multi-person conversation and passes that context to an AI agent.
Status: working prototype.
What I learned
The value isn’t in one smart system. It’s in linking systems, so each one extends what the others can do.