Description
Amazon is hiring a Software Development Engineer to design and implement the inference data plane for large language models on custom hardware. The role focuses on developing and optimizing compute kernels, validating LLM architectures, integrating custom accelerator backends with serving frameworks, building test and profiling infrastructure, and driving latency and throughput improvements across the stack. It requires a bachelor's degree or equivalent, at least four years of full software development experience, strong C/C++ proficiency, and experience with GPUs, DSPs, or custom accelerators.
