Build real-time AI agent infrastructure: Design and operate the stateful, low-latency runtime that powers voice and chat AI agents — from LLM streaming and conversation state management to graceful recovery and multi-channel support Solve distributed systems problems: Own session management across s