Principal Software Engineer – Robot Applications & Voice AI

A Moving Experience. Principal Software Engineer – Robot Applications & Voice AI Location : California Bay area Humanoid robots become truly useful when they understand human intent and convert it into meaningful action. At Cerence , we are building the cognitive and application layer that combines Voice AI, Agentic AI, multimodal reasoning, and robotic capabilities into intuitive human-robot experiences. Role Overview The Principal Software Engineer drives the application and cognitive layer of our humanoid robot. Working directly with the core robotics platform designed by our Architect, you will turn human intent into high-level physical actions. This role focuses on building advanced voice applications, multi-modal context systems, and programming the core business logic and behavior trees that make the robot interactive, helpful, and intuitive. Your Impact Build software powering the next generation of intelligent humanoid robots. Work on Voice AI, Agentic AI, LLMs, VLMs, and Physical AI products. Collaborate with experts in robotics, conversational AI, embedded systems, and cloud computing. See your work deployed in real-world robotic environments. Key Responsibilities Application Layer & Business Logic: Implement high-level application software, interaction workflows, and robotic behavioral state machines. Intent-to-Action Orchestration: Build the core logic that translates human speech, gestures, environmental cues, and contextual information into deterministic robotic tasks. Multi-Modal Interaction Fusion: Integrate vision pipelines, object detection, scene understanding, spatial tracking, and voice interactions to provide contextual awareness. Voice Application Engineering: Develop robust voice applications managing STT, NLU, dialog management, wake-word detection, and TTS interfaces. Agentic AI Development: Build orchestration workflows combining LLMs, VLMs, memory systems, tool calling, reasoning engines, and robotic capabilities. Robot Capability Integr