Design and implement core software for advanced in-vehicle voice assistants that leverage both cloud and edge-based large language models (LLMs). This role focuses on building intelligent middleware that orchestrates speech recognition, natural language understanding, and text-to-speech components into a responsive, low-latency user experience.
Key Responsibilities
- Collaborate with LLM development teams to integrate software development kits into embedded automotive systems
- Develop and maintain communication layers that coordinate ASR, NLU, and TTS subsystems
- Architect hybrid inference solutions that dynamically route requests between cloud and on-device LLMs
- Reduce end-to-end system latency through optimization of processing pipelines
- Manage resource efficiency, including CPU utilization, memory footprint, and cold-start performance
Required Expertise
- Minimum of three years of professional software development experience
- Strong proficiency in C++11/14 for embedded systems
- Hands-on experience with Linux and Android Automotive OS environments
- Familiarity with LLMs, agent architectures, retrieval-augmented generation (RAG), and inference frameworks
- Understanding of inter-thread and inter-process communication mechanisms
Technology Environment
Work within a stack centered on C++11/14, Linux, and Android Automotive OS, integrating speech technologies and LLM runtimes. Focus on secure, efficient, and scalable solutions for next-generation vehicle interfaces.
Work Environment
This is a local position requiring on-site presence. The team values security awareness, technical precision, and inclusive collaboration across global engineering groups.
