Senior Research Engineer

• Investigate and mitigate performance bottlenecks in large-scale distributed training and inference systems.
• Develop and implement both low-level (operator/kernel) and high-level (system/architecture) optimization strategies.
• Translate research models and prototypes into highly optimized, production-ready inference systems.
• Explore and integrate inference compilers such as TensorRT, ONNX Runtime, AWS Neuron and Inferentia, or similar technologies.
• Design, test, and deploy scalable solutions for parallel and distributed workloads on heterogeneous hardware.
• Facilitate knowledge transfer and bidirectional support between Research and Engineering teams, ensuring alignment of priorities and solutions.

• Strong expertise in the Python ecosystem and major ML frameworks (PyTorch, JAX).
• Experience with lower-level programming (C++ or Rust preferred).
• Deep understanding of GPU acceleration (CUDA, profiling, kernel-level optimization); TPU experience is a strong plus.
• Proven ability to accelerate deep learning workloads using compiler frameworks, graph optimizations, and parallelization strategies.
• Solid understanding of the deep learning lifecycle: model design, large-scale training, data processing pipelines, and inference deployment.
• Strong debugging, profiling, and optimization skills in large-scale distributed environments.
• Excellent communication and collaboration skills, with the ability to clearly prioritize and articulate impact-driven technical solutions.

Pay range: $240K - $275K