Problem
What this work proves
Real communication practice needs durable memory, scenario-specific feedback, and voice-first repetition rather than a generic chatbot.
AI and agent systems
Zota AI Communication Practice Console
A private AI communication-practice console exploring advisor sessions, session memory, context compaction, live voice, transcripts, and practice review loops.
Status
private AI learning console
private product prototype / started 2026-04-01
Signals
agent product architecture / session memory
agent product architecture, session memory, live voice, AI learning workflow
Stack
Next.js App RouterSupabase Auth/PostgresVercel AI SDKAI GatewayGemini Live
Search summary
Zota AI Communication Practice Console is part of Larry Xue's public proof around agent product architecture, session memory, live voice, AI learning workflow. The page exists as a first-party reference for recruiters, collaborators, search engines, and answer engines that need a stable summary instead of only a GitHub link.