Applied Research Scientist, World Model

What you'll do

• Translate shopper behavior into structured demand representations and estimate willingness-to-pay, home-level demand curves, and market-clearing dynamics.
• Develop latent state-space, multi-tower, or sequential models with calibrated uncertainty; evaluate forecasts, rollouts, and causal validity.
• Lead open-ended research from scoping through production impact, collaborating across teams and mentoring scientists.

What they ask for

• PhD in Computer Science, Machine Learning, Statistics, Economics, or a related quantitative field.
• Hands-on experience building latent state-space or world models, including belief-state representations, stochastic transitions with calibrated uncertainty, and multi-step rollout evaluation.
• Experience deriving demand representations from large-scale user behavior while correcting for selection bias, plus practical causal inference in observational data.
• Strong programming skills, especially Python, and experience with PyTorch or TensorFlow; proven ability to lead ambiguous research projects.

Nice to have

• Economics fluency in market microstructure, search-and-matching, or structural/discrete-choice demand models; experience with agent-based or market simulation or simulation-based reinforcement learning.
• Published work in world models, model-based RL, causal inference, or structural demand estimation; multimodal learning, bitemporal data, or marketplace experience.

Pay

$195,600.00 - $312,400.00 annually (base salary); $185,800.00 - $296,800.00 annually (base salary)

Tools

Python, PyTorch, TensorFlow, RSSM/Dreamer-family, JEPA-style latent prediction, event-sequence transformers

AI focus

Develop a housing-market World Model for multi-step forecasting and forward/backward planning, grounded in shopper behavior, economic quantities, causal inference, and calibrated uncertainty.