Senior Research Scientist, Nemotron Post-training
Core
Designing and scaling post-training pipelines for open-source generative AI foundation models (Nemotron), focusing on agentic reinforcement learning and large-scale model deployment.
Role type
Senior Research Scientist / Engineer (Post-training & RL)
Builds
Next-generation post-training technologies, training data, benchmarks, and software (NeMo-RL, Nemo-Gym) for open-source foundation models.
Domain
Generative AI, Reinforcement Learning, Large Language Models
Deliverable
production ML models
Required skills
Reinforcement learning, Agentic systems, Data curation, Model training, Inference optimization, Large-scale training orchestration
Preferred skills
Real-world traffic feedback optimization, Leading foundation model RL experience
Technologies
vLLM, SGLang, TRT-LLM, NeMo-RL, Nemo-Gym
Responsibilities
Develop training data and benchmarks for agentic RL; Implement data and training infrastructure; Collaborate on vendor data acquisition; Solve end-to-end post-training challenges from orchestration to deployment; Publish research at academic and industry conferences.
Seniority
Senior, hands-on IC
