Principal Applied Scientist, Generative AI

This position is listed on behalf of a partner company, which manages all applications and next steps. Our partner is looking for a Principal Applied Scientist, Generative AI based in the United States. As Principal Applied Scientist, Generative AI, you will define the scientific direction for advanced large language models powering next-generation AI-driven home-shopping experiences. You will lead and mentor a team of applied scientists while remaining deeply involved in research, experimentation, and technical execution. Your work will focus on LLM post-training, reward modeling, and evaluation to improve the quality, reliability, and effectiveness of agentic AI systems. You will translate cutting-edge advances in generative AI and reinforcement learning into practical innovations that serve millions of users. Working in a collaborative, research-driven environment, you will partner with Product, Engineering, Platform, and Data teams to turn ambitious ideas into measurable product outcomes. This is an opportunity to influence AI strategy, solve complex scientific challenges, and shape intelligent consumer experiences at significant scale.

Accountabilities

• Define AI research strategy: Establish the scientific roadmap for LLM post-training, reward modeling, and evaluation of agentic systems, identifying high-impact opportunities to advance model capabilities.
• Lead and mentor scientific talent: Provide technical direction, mentorship, and research guidance to a team of applied scientists, fostering scientific rigor, innovation, and continuous learning.
• Drive end-to-end research initiatives: Independently tackle complex, ambiguous problems, translating product vision into research hypotheses, experimental designs, model improvements, and the data required to achieve meaningful results.
• Advance LLM performance: Develop and refine post-training approaches, including supervised fine-tuning, preference optimization, reinforcement learning from human feedback, and related techniques, to improve model quality and behavior.
• Build effective evaluation frameworks: Establish rigorous methods for measuring the performance, reliability, and effectiveness of language models and agentic AI systems, using experimental findings to guide further development.
• Translate research into production: Apply state-of-the-art generative AI research to practical product improvements, ensuring scientific advances deliver tangible value in consumer-facing experiences.
• Collaborate across disciplines: Partner with Product, Engineering, Platform, and Data teams to align research priorities with business objectives and successfully integrate AI capabilities into production systems.
• Communicate scientific insights: Present research findings, technical trade-offs, strategic recommendations, and long-term opportunities clearly to executive leadership and cross-functional stakeholders.
• Shape technical excellence: Promote reproducible experimentation, sound research practices, and evidence-based decision-making while helping establish the team’s long-term technical direction.

Requirements

• Education: PhD in Computer Science, Machine Learning, or a related discipline, or a master’s degree supported by equivalent research experience.
• Applied science expertise: Deep, hands-on experience with large language models or foundation models, particularly in post-training techniques such as supervised fine-tuning (SFT), Direct Preference Optimization (DPO), reinforcement learning from human feedback (RLHF), reinforcement fine-tuning (RFT), or Group Relative Policy Optimization (GRPO).
• Reward modeling and evaluation: Strong understanding of reward modeling, preference learning, model evaluation, and experimental methodologies for improving LLM performance.
• Technical leadership: Demonstrated ability to lead and develop other scientists or engineers while remaining technically hands-on, or to lead a research group with a strong record of scientific contributions and publications.
• Research execution: Proven ability to independently define research problems, design experiments, analyze results, and turn ambiguous challenges into actionable technical solutions.
• Programming and machine learning: Proficiency in Python and machine learning frameworks such as PyTorch, including the ability to read, debug, review, and guide experimental code.
• Strategic influence: Strong communication and collaboration skills, with the abil…