Director, Research - Human Data Systems

Snorkel AI came out of the Stanford AI lab and is built on the principle that data development is the core of AI. For years, we have pioneered data-centric methods that power the world’s most sophisticated AI implementations from top enterprises to frontier AI labs. We are looking for a Director of Research to lead a team focused on data-centric research.

About the Role

You will be responsible for developing cutting-edge human-in-the-loop systems that move beyond labeling to guiding experts on complex, multi-step tasks. From rapidly prototyping and testing methods with 1000s of users to innovating core data-centric methods, you will have the opportunity to lead how the data that powers the next generation of frontier models is created and curated.

Main Responsibilities

• Team Building & Leadership: Recruit for and lead a team of applied researchers, setting the team’s culture and setting a high standard for quality and velocity
• The Research-to-Product Bridge: Act as the primary link between Research, Product, and Engineering. You will oversee the development of experimental POCs to improve workflow efficiencies and ensure these innovations are successfully hardened and integrated into the platform
• Strategic Roadmapping: Anticipate the next bottlenecks in the AI data (e.g., automated RLHF, agentic evaluation, or specialized model alignment) and pivot the team to develop the programmatic solutions that address them
• Operational Excellence: Partner with the Operations team to validate research-driven workflows in real-world production environments, ensuring they scale across diverse and complex domain-specific tasks.

 

What We’re Looking For

• 5+ years of experience in applied AI, research or machine learning roles, with at least 4+ years of experience managing technical teams
• A player coach who is willing to be hands-on with coding and research
• A proven track record of thriving in fast-paced, ambiguous environments, comfortable managing competing priorities across cross-functional stakeholders
• Enjoys solving real-world problems in the data space and bridging research and production
• Strong, practical experience working with LLMs, agentic workflow
• Bonus: Experience with data annotation workflows and/or building internal tooling for them