Applied Research Scientist, World Model
Core
Build a foundational 'World Model' of the housing market that ties together supply, demand, and macro conditions to answer multi-step 'what if' questions for Zillow's AI Mode.
Role type
Principal Applied Scientist (World Models & Causal Inference)
Builds
A single, uncertainty-aware model of the housing market serving Zillow's AI Mode (Voyager).
Domain
Real Estate / Housing Market / AI
Deliverable
production ML models
Required skills
Latent state-space modeling, Causal inference, Willingness-to-pay estimation, Multi-step rollout planning, Python, PyTorch/TensorFlow
Preferred skills
Economics fluency, Agent-based simulation, Model-based RL, Multimodal representation learning
Technologies
PyTorch, TensorFlow, Python
Responsibilities
Build the demand layer of the World Model; Design and train latent state-space and sequential dynamics models; Estimate willingness-to-pay and market clearing dynamics; Apply causal identification in observational data; Contribute to the World Model evaluation battery; Drive ambiguous research end-to-end.
Seniority
Principal, hands-on IC with technical leadership
