Principal Applied Scientist (User Modeling)

📋 Description

• Develop advanced user models that help AI agents understand user journeys, infer intentions
• Lead end-to-end large language model post-training pipelines, including SFT, DPO, and RFT/GRPO
• Develop and train efficient, multi-category process reward models (PRMs) that generate scalar
• Design on-policy and online assessment frameworks that score live agent trajectories using
• Convert offline evaluation rubrics into generalizable reward functions for assessing production
• Apply causal inference and reinforcement learning techniques, including sequential decision

🎯 Requirements

• PhD in Computer Science, Statistics, Economics, Machine Learning, or related quantitative field
• Hands-on experience modeling user behavior at scale, including sequence models on event streams
• Practical expertise in causal inference methods or reinforcement learning for real-world sequential
• Experience with LLM post-training methodologies, reward modeling, preference optimization, or
• Working proficiency in Python and machine learning frameworks such as PyTorch or TensorFlow.
• Demonstrated ability to independently scope and solve ambiguous technical problems, design rigorous

🎁 Benefits

• Competitive annual base salary: $195,600–$312,400 in California, Connecticut, Maryland
• Equity compensation eligibility.
• Remote work flexibility for eligible US-based employees.
• High-impact research opportunities on advanced AI systems.
• Access to cutting-edge technology in reinforcement learning, causal inference, LLM post-training
• Professional growth opportunities including influence on scientific strategy and mentoring.