Software Engineer, ML Performance

The ML Performance Engineer will own the full lifecycle of ML model deployment on robots, from handoff to system integration, and will be responsible for converting, optimizing, and integrating trained models for Jetson platforms using NVIDIA tools. This role involves developing and optimizing CUDA kernels and pipelines for low-latency, high-throughput model inference, profiling and benchmarking ML workloads, and identifying and removing compute and memory bottlenecks. The engineer will also design and implement strategies for model quantization and compression, ensure model robustness under resource constraints, manage memory layout and concurrency for optimal GPU/CPU usage, build benchmarking pipelines, collaborate with QA and systems teams for validation, and work closely with ML researchers to influence model architectures for edge deployability.