AI Agent 应用工程师（语音与新型人机交互方向）
Core
Architecting autonomous AI Agent systems that bridge large language models with physical hardware to enable next-generation human-computer interaction (HCI) via voice and natural language.
Role type
Senior IC AI Agent Architect (Voice & HCI)
Builds
Stateful multi-agent orchestration systems, real-time voice-to-action prototypes, and unified hardware-software context protocols.
Domain
Consumer electronics / AI Agents / Voice Engineering
Deliverable
production ML models | product features
Required skills
Python, C/C++/Go/Rust, Agent system design, ReAct/Plan-and-Solve/Reflection paradigms, LangGraph/Temporal/CrewAI, real-time audio streaming (WebSockets/WebRTC), STT/TTS engineering, vector/graph databases, MCP protocol, OS-level APIs (Windows/macOS), hardware communication protocols.
Preferred skills
IoT/robotics/hardware integration projects, high-star GitHub repos on multi-agent/MCP, hackathon awards.
Responsibilities
Translate user interaction pain points into executable Agent decision logic and tool chains; build hardware-to-voice-to-automation prototypes; design robust stateful multi-agent systems for complex task orchestration; implement long-term memory architectures with dynamic fusion of episodic and procedural memory; unify hardware status and cloud LLM contexts via MCP; design deterministic safety boundaries and human-in-the-loop confirmation mechanisms.
Seniority
Senior, hands-on IC