Senior/Staff Software Engineer - Machine Learning & System Optimization

What you'll do

• Allocate CPU, GPU, and interconnect resources among on-vehicle models and inference engines; improve utilization through model sharing, fusion, and scheduling.
• Optimize large-scale models using quantization, pruning, mixed precision, and parameter-efficient fine-tuning; build TensorRT conversion and compilation pipelines.
• Develop production-level C++ and CUDA code, including custom kernels, for low-latency, real-time inference on vehicle systems.

What they ask for

• Deep experience optimizing CPU/GPU systems for low latency or high throughput, including real-time constraints.
• Deep expertise in model quantization (PTQ, QAT) and mixed-precision inference.
• Proficiency developing and optimizing custom ML operations and TensorRT plugins with CUDA kernels.
• Production-level C++ (14/17/20) and Python skills, including concurrent, memory-safe inference code for edge devices.

Nice to have

• Experience in high-performance robotics applications such as AVs, drones, or robots.
• Familiarity with autonomous-driving perception, multimodal sensor processing, foundation models, or edge deployment technologies such as TensorRT-LLM.

Pay

USD 254,000–350,000 per-year-salary; base salary

Tools

TensorRT, C++ (14/17/20), Python, CUDA, LoRA, QLoRA, TensorRT-LLM, INT8

AI focus

Compresses, accelerates, schedules, and deploys multimodal perception models, LLMs, VLMs, and foundation models for real-time vehicle inference.