As a Principal Applied Scientist at Zillow, you will lead the design and deployment of reinforcement learning (RL) post-training systems for large-scale AI models, ensuring they align with user value, safety, and business objectives. You will develop and implement pipelines that fine-tune models using supervised learning, preference modeling, and RL-based alignment techniques like RLHF and DPO to optimize multi-objective criteria such as helpfulness and compliance. Your role involves creating reward models from conversational logs and behavioral signals for post-training and reinforcement learning, improving the efficiency of training and evaluation processes with off-policy methods and controlled rollouts. You will also mentor team members and contribute to Zillow’s broader AI roadmap through thought leadership and guidance, working in a fast-moving environment that values experimentation and continuous improvement.
Lead the design and deployment of RL post-training systems for production models.
Design and implement post-training pipelines combining supervised fine-tuning and RL-based alignment approaches.
Develop reward models balancing constraints like safety, fairness, and customer satisfaction using human/AI feedback.
Translate conversational logs into training signals for post-training and reinforcement learning.
Mentor applied scientists and engineers in RL, post-training, and evaluation techniques.
Improve efficiency of training and evaluation through off-policy evaluation and controlled rollouts.
What we're looking for
PhD or equivalent experience in Computer Science, Electrical Engineering, Statistics, or related field.
Strong expertise in reinforcement learning and post-training techniques like DPO, RLHF/RLAIF, preference modeling.
Experience designing and implementing reward models for multi-objective optimization with constraints on safety, fairness, compliance.
Proven track record of working with cross-functional teams in high-stakes domains such as finance or healthcare.
Demonstrated ability to mentor applied scientists and engineers, raising the technical bar in RL and evaluation.
Strong background in modern transformer-based models and tooling including LLMs, multimodal models, vector search.
Market check
Salary context
Above market
How this pay compares to similar roles
Similar $212k
This role $248k
$161k
most similar roles pay here
$321k
This role pays more than
79%
of similar roles. Most pay
$176,937–$246,150
— the shaded band above.
At the midpoint, this role pays about
$248k
versus about
$212k
for comparable roles.
Based on 240 similar postings.
Employer
About Zillow
Zillow Group is a leading real estate and rental marketplace providing consumers with data, tools, and services to find, buy, sell, rent, and finance homes, and connecting buyers with agents and lenders. Industry: Real Estate Technology & Marketplace
Zillow currently has
35 open roles
on FindRole.
Listed pay typically runs
$160,900–$257,100
across 35 roles with salary data.