A fully local voice‑controlled AI agent uses Whisper for speech‑to‑text, LLaMA for language understanding, and LangGraph for tool execution. It accepts microphone or audio files, transcribes commands, detects intent, and creates files, writes code, or summarizes text.
Comments