Principal Applied Scientist, Agentic AI

About the team

The Agentic AI team at Zillow is at the forefront of transforming the real estate industry by helping millions of people use AI technologies to find their next home. We aim to redefine the home shopping and transaction experience by providing an always-on, multimodal assistant that integrates deep user and home insights, along with sophisticated reasoning. By embedding domain knowledge and specialized expertise in these assistants, we empower home buyers and real estate professionals to gain efficiency while maintaining Zillow's competitive edge.

This lean, customer-focused team comprises scientists and engineers working together to deliver delightful customer experiences. We strongly believe in accountability, collaboration, and transparency.

About the role

Zillow is seeking a Principal Applied Scientist on the Agentic AI team to research and develop advanced, multimodal foundational technologies (CV, NLP, 3D) for next-generation understanding of homes, both indoors and outdoors. They will be hands-on in prototyping ideas, evaluating, and deploying models to Zillow's highly scalable infrastructure platforms.

As a Principal Applied Scientist in the Agentic AI team, you will:

• Design, prototype, and develop multi-agent image-forward, voice-assisted experiences.

• Improve real estate shopping experience using computer vision tasks such as VLM captioning, object detection and localization, segmentation, and multimodal semantic retrieval

• Serve as the focal point for developing a science roadmap, driving alignment across stakeholders on technical capabilities, timelines, and prioritization

• Represent the team at conferences, publish in top-tier journals, like CVPR, NeurIPS, ICCV, ACL, and SIGGRAPH, and champion internal best practices

• Build innovative solutions that personalize using user behavior, context signals, and real-time environmental data

• Distill complex research findings and system designs into actionable insights for diverse audiences

This role has been categorized as a Remote position. “Remote” employees do not have a permanent corporate office workplace and, instead, work from a physical location of their choice, which must be identified to the Company. U.S. employees may live in any of the 50 United States, with limited  California, Connecticut, Maryland, Massachusetts, New Jersey, New York, Washington state, and Washington DC the standard base pay range for this role is $ - $ annually.

This base pay range is specific to these locations and may not be applicable to other  Colorado, Hawaii, Illinois, Minnesota, Nevada, Ohio, Rhode Island, and Vermont the standard base pay range for this role is $ - $ annually. The base pay range is specific to these locations and may not be applicable to other  addition to a competitive base salary this position is also eligible for equity awards based on factors such as experience, performance and location.

Actual amounts will vary depending on experience, performance and location. Employees in this role will not be paid below the salary threshold for exempt employees in the state where they reside.

Who you are

As a Principal Applied Scientist, you are a roll-up-the-sleeves and get-it-done researcher who can marry scientific rigor with large-scale engineering. We are looking for someone who has:

• Ph.D. or equivalent experience in Computer Vision, NLP, or related fields, with emphasis on multimodal learning

• Deep expertise in developing and scaling perception models that combine imagery, spatial data, and language

• Proven science leadership in defining a science roadmap for a team and mentoring other scientists

• Track record of publishing high-impact research in top venues like CVPR, NeurIPS, ICCV, ACL, and SIGGRAPH

• Experience developing multi-agent dialogue systems with long context multimodal chain of thought reasoning.

• Hands-on experience fine-tuning voice-AI models for automatic speech recognition, turn detection, and speech generation is a bonus

• Experience with geospatial datasets and 3-D vision techniques, like point clouds, Gaussian splatting, or neural radiance fields, is a plus

• Track record of shipping large-scale, high-impact consumer-facing solutions…