Senior Deep Learning Scientist, Multimodal Agentic RL
Core
Develop, train, fine-tune, and deploy large language models for agentic systems encompassing audio-visual reasoning, tool usage, and document understanding.
Role type
Senior Deep Learning Scientist (Multimodal Agentic RL)
Builds
Multimodal foundation models (Nemotron Omni, VoiceChat) for reasoning, planning, and acting across diverse modalities.
Domain
Artificial Intelligence / Deep Learning / Reinforcement Learning
Deliverable
production ML models
Required skills
Deep learning, Reinforcement learning, Applied mathematics, Python, Scalable model development, Transformers, Mixture-of-experts, Post-training methods (RLHF/RLVR/MOPD), Dataset management
Preferred skills
Top-tier AI publications, Large-scale distributed infrastructure training, Deep reinforcement learning in simulation/gaming, Audio/speech AI, Embodied AI systems
Technologies
PyTorch, Nemotron Omni, VoiceChat
Responsibilities
Develop and deploy LLMs for agentic systems; Advance post-training and alignment methods; Research agentic reasoning and grounded perception; Lead collection and benchmarking of multimodal datasets.
Seniority
Senior, hands-on IC
