Research Intern – AI Incubation

Zoom • Seattle, WA

3d • Hybrid

About The Position

What you can expect Zoom’s AI Incubation team is looking for a PhD Research Intern to dive into the next wave of LLM innovation. You will work alongside a world-class group of PhDs and applied scientists, contributing to high-impact projects in post-training, reinforcement learning (RL), and federated AI. As an intern, you won't just be watching from the sidelines. You will own a specific research project, conduct experiments at scale, and help develop the breakthroughs that power the next generation of the Zoom AI Companion. This is a unique opportunity to see how frontier AI research is translated into a product used by millions. About the team The AI Incubation team is a high-impact applied research group known for building one of the industry's best-performing federated AI systems. We operate at the frontier of: Agentic Intelligence: Moving beyond chat to models that "do." Federated AI: Privacy-preserving, edge-to-cloud learning across diverse model ecosystems (Anthropic, OpenAI, Google). Advanced Alignment: Pushing RLHF, DPO, and RLAIF to new heights of reasoning and reliability.

Requirements

• PhD Candidate Currently enrolled in a PhD program in Computer Science, Machine Learning, Artificial Intelligence, or a related quantitative field.
• Strong Technical Foundation Solid programming skills with hands-on experience in PyTorch.
• Relevant Research Experience Experience in at least one of the following areas: LLM post-training, including SFT or RLHF Federated learning Multimodal modeling Agentic or workflow-based AI systems
• Research Mindset Proven ability to solve open-ended research problems, demonstrated through: Publications at top-tier conferences (for example NeurIPS, ICML, ICLR) Meaningful open-source contributions Advanced research projects or impactful academic work
• Collaboration and Communication Able to clearly explain complex technical ideas. Comfortable working in a fast-paced, iterative research environment and collaborating across teams.

Nice To Haves

• Bonus points for experience with modern LLM tooling such as Hugging Face Transformers, DeepSpeed, or FlashAttention.

Responsibilities

• Execute Research Projects: Design and implement experiments in areas like LLM fine-tuning, preference optimization (DPO/PPO), or distributed federated learning.
• Prototype & Evaluate: Build and benchmark new model architectures or training recipes to improve reasoning, personalization, and safety.
• Collaborate: Work closely with senior scientists to refine research hypotheses and troubleshoot large-scale training runs.
• Document & Present: Synthesize your findings into internal reports or potential publications, presenting your work to the broader AI organization.
• Stay Curious: Keep pace with the latest ArXiv drops and open-source developments to ensure our methods remain state-of-the-art.

What This Job Offers

• Career Level: Intern
• Education Level: Ph.D. or professional degree