What you can expect
Zoom's AI Incubation team is seeking a Research Scientist to contribute to the next wave of LLM post-training, reinforcement learning, and federated AI innovation. You will join a world-class group of PhDs and applied scientists focused on pushing the boundaries of model alignment, reasoning, and agentic intelligence.
As a Research Scientist, you will advance Zoom's applied AI research, working with cutting-edge LLM post-training, RLHF/DPO/RLAIF, federated AI, and multimodal intelligence to deliver breakthroughs that redefine human-AI collaboration. You will participate in incubation efforts that transform bold research ideas into scalable, production-ready systems powering the next generation of Zoom AI Companion.
About the Team
The AI Incubation team is a high-impact applied research group that has built one of the best-performing federated AI systems in the industry. The team operates at the frontier of LLM post-training, reinforcement learning for agentic AI, multimodal understanding, federated AI, and AI evaluation.
Zoom's federated AI architecture—spanning Anthropic, OpenAI, and Google models—enables privacy-preserving, edge-to-cloud learning that adapts to user context while maintaining enterprise-grade governance. The Incubation team plays a central role in advancing this federated AI leadership, developing new methods for distributed model optimization, personalized prefix learning, and secure collaboration.
Responsibilities
- Conducting frontier research in LLM post-training, reinforcement learning, and federated AI to achieve state-of-the-art reasoning, personalization, and reliability.
- Developing and implementing evaluation and benchmarking frameworks for model performance, safety, and user experience across distributed environments.
- Contributing to model architectures, fine-tuning strategies, and reinforcement learning pipelines across Zoom's federated AI ecosystem.
- Partnering with product and infra teams to translate research prototypes into scalable, production-grade AI systems that leverage edge and cloud collaboration.
- Participating in a culture of scientific rigor, creativity, and rapid experimentation.
- Exploring emerging AI paradigms—from agentic reasoning and multimodal fusion to self-improving federated systems—to continuously advance Zoom's AI frontier.
What we’re looking for
- Have a PhD or advanced degree in Computer Science, Artificial Intelligence, Machine Learning, or a related field.
- Demonstrate proven expertise in LLM post-training, RLHF/DPO/PPO, federated learning, or reinforcement learning for reasoning and alignment.
- Have deep understanding of large-scale distributed training systems and experience with PyTorch, Transformers, DeepSpeed, or CUDA.
- Hold a publication or open-source record demonstrating innovation in LLM optimization, federated AI, or agentic intelligence.
- Demonstrate passion for applied AI research that bridges scientific discovery and real-world impact.
- Have experience implementing AI research concepts and translating them into practical applications.