About the job
Join NVIDIA and help build the software that will define the future of generative AI. We are looking for a research engineer who is passionate about open-source and excited to create our next-generation post-training software stack. You will work at the intersection of research and engineering, collaborating with the Post-Training and Frameworks teams to invent, implement, and scale the core technologies behind our Nemotron models.
Responsibilities
Work with applied researchers to design, implement and test next generation of RL and pos-training algorithms
Contribute and advance open source by developing NeMo-RL, Megatron Core, and NeMo Framework and yet to be announced software
You will be engaged as part of one team during Nemotron models post-training
Solve large-scale, end-to-end AI training and inference challenges, spanning the full model lifecycle from initial orchestration, data pre-processing, running of model training and tuning, to model deployment.
Work at the intersection of computer-architecture, libraries, frameworks, AI applications and the entire software stack.
Performance tuning and optimizations, model training with mixed precision recipes on next-gen NVIDIA GPU architectures.
Publish and present your results at academic and industry conferences
Qualifications
Minimum
BS, MS or PhD in Computer Science, AI, Applied Math, or related fields or equivalent experience
6+ years of proven experience in machine learning, systems, distributed computing, or large-scale model training.
Experience with AI Frameworks such as Pytorch or JAX
Experience with at least one inference and deployment environments such as vLLM, SGLang or TRT-LLM
Proficient in Python programming, software design, debugging, performance analysis, test design and documentation.
Strong understanding of AI/Deep-Learning fundamentals and their practical applications.
Preferred
Contributions to open source deep learning libraries
Hands-on experience in large-scale AI training, with a deep understanding of core compute system concepts (such as latency/throughput bottlenecks, pipelining, and multiprocessing) and demonstrated excellence in related performance analysis and tuning.
Expertise in distributed computing, model parallelism, and mixed precision training
Prior experience with Generative AI techniques applied to LLM and Multi-Modal learning (Text, Image, and Video).
Knowledge of GPU/CPU architecture and related numerical software.