Principal Applied Scientist, Generative AI

Zillow

3d•$185,800 - $312,400•Remote

About The Position

The Principal Applied Scientist will set the scientific direction for the large language models behind Zillow’s AI assistant and lead the scientists who build them. The role defines the research agenda for LLM post-training, reward modeling, and agentic-system evaluation; provides technical direction and mentorship to a group of applied scientists; and drives ambiguous, high-impact problems from product vision through experiments, models, and data. The scientist translates generative AI research into practical improvements for a product used by millions of home shoppers and communicates scientific strategy and results to executives and cross-functional partners. The position is part of Zillow’s Agentic Foundations team, a focused group of 15–20 applied scientists within the broader Agentic AI organization. The team works in a flat structure, with scientists owning areas and setting direction while partnering with product, engineering, platform, and data teams. Its work supports agentic skills across the home-shopping journey, including search, personalized guidance, offer strategy, and financing, at a scale of hundreds of millions of requests per day. This role is categorized as remote. Remote employees do not have a permanent corporate office and work from a physical location of their choice that must be identified to the Company. U.S. employees may live in any of the 50 states, with limited exceptions. The role offers a competitive base salary and is eligible for equity awards. Salary ranges vary by location.

Requirements

PhD in Computer Science, Machine Learning, or a related field, or a master's degree with equivalent research experience.
Deep, hands-on expertise in large language models or foundation models, including post-training (SFT, DPO, RLHF, RFT/GRPO), reward modeling, or LLM evaluation.
A track record of leading others, either by managing teams of scientists or engineers while remaining technically hands-on, or by leading a research group with a strong publication record.
Ability to independently scope and solve ambiguous problems and influence technical direction across teams.
Proficiency in Python and ML frameworks such as PyTorch, including the ability to read, debug, and guide experimental code.

Nice To Haves

Publications at top venues such as NeurIPS, ICML, ICLR, ACL, or EMNLP.
Experience applying LLMs in consumer-facing products at scale.
Experience with agentic AI systems and their evaluation.
Background in reinforcement learning or preference learning.

Responsibilities

Define the research agenda for LLM post-training, reward modeling, and evaluation of agentic systems.
Lead and grow a group of applied scientists, providing technical direction, mentorship, and a high bar for scientific rigor.
Drive ambiguous, high-impact problems end to end, from product vision through experiments, models, and data.
Translate state-of-the-art generative AI research into practical improvements in a product used by millions of home shoppers.
Communicate scientific strategy and results to executives and cross-functional partners.