[Remote] Senior Research Engineer - Enterprise Products at NVIDIA

The Senior Research Engineer will develop new models and algorithms focused on Large Language Models, Natural Language Processing, and Deep Learning. Responsibilities include designing and implementing multi-node serving architectures for disaggregated serving and distributed LLM inference, optimizing multi-LoRA and other PEFT technique inference serving systems, and applying sophisticated quantization techniques (FP4/INT4, FP8) to reduce model footprint while preserving quality. The role involves implementing speculative decoding (draft target, eagle, medusa etc.) and other latency optimization strategies. Additionally, the engineer will demonstrate good engineering practices, mentor other team members, and collaborate with engineering teams across NVIDIA to ensure seamless software integration within the NVIDIA accelerated serving stack. The position also involves conceptualization, applied research, engineering for optimized inference, and deployment, contributing to all steps of the machine learning lifecycle and collaborating with research teams, engineers, and the open-source community.