Senior Research Scientist - Monetization

About Us

Yahoo Mail is the ultimate consumer inbox. The Mail Intelligence platform builds next-generation platforms and services that deliver deeply personalized content to hundreds of millions of users across devices. We process billions of messages and data at petabyte scale and use cutting‑edge machine learning to power monetization, recommendations, and personalization.

• You are an expert in developing and applying state‑of‑the‑art machine learning and deep learning models to solve complex, high‑stakes monetization and personalization problems.
• You thrive in a research‑oriented environment, pushing the boundaries of innovation while simplifying model architectures and delivering measurable business impact.
• You have a strong academic foundation and a demonstrated commitment to personalization, user modeling, and monetization strategies for large‑scale systems.
• You excel in translating theoretical research into practical, high‑impact algorithmic applications—particularly in ranking, recommendation systems, ads targeting, causal uplift, pricing strategy, and reinforcement learning.
• You are a hands‑on research leader with deep expertise in designing, training, and evaluating large‑scale models, including transformer‑based and deep learning architectures.
• You possess an AI‑forward mindset, actively seeking to optimize your research and modeling workflow through modern AI diagnostics, prompt engineering, and automated evaluation frameworks.
• You have a collaborative, consensus‑building mindset, bringing excellent communication skills to resolve complex theoretical tradeoffs and guide cross‑functional squads.

• Lead the research and algorithmic design of advanced models for ranking, personalization, recommendation systems, ads targeting, causal uplift modeling, dynamic pricing, and revenue forecasting.
• Design, train, evaluate, and prototype state‑of‑the‑art machine learning, deep learning, and reinforcement learning models, taking full ownership of experimental integrity and long‑term scientific risks.
• Establish robust experimental standards and guide scalable modeling workflows on platforms like GCP to streamline model training, validation, and evaluation.
• Incorporate LLM‑driven synthesis into monetization models to prototype hybrid architectures and accelerate relevance tuning.
• Collaborate with engineering partners to transition machine learning models from prototype to production, making strategic tradeoffs between theoretical complexity and operational constraints.
• Proactively simplify existing model architectures, address core modeling bottlenecks, and establish rigorous scientific standards across multiple projects.
• Leverage AI‑assisted coding and prototyping tools to accelerate model development and evaluation.
• Stay abreast of the latest research trends, contributing to Yahoo's engineering and scientific brand through publications, patents, or technical presentations.
• Mentor, guide, and conduct technical assessments for senior and mid‑level research scientists and machine learning engineers, fostering a culture of continuous learning and AI experimentation.

• PhD (preferred) or Master's degree in Computer Science, Statistics, Applied Mathematics, or a related field.
• 8+ years of industry experience (or 5+ years with a PhD) in machine learning, deep learning, or related fields, with a track record of designing and scaling advanced ranking, personalization, ads targeting, or revenue optimization systems.
• Strong fundamentals in machine learning and deep learning, with proven expertise in optimizing business metrics (e.g., CTR, CVR, eCPM) alongside model‑centric loss functions.
• Comprehensive understanding of deep learning architectures, including transformer‑based models, and hands‑on experience validating model outputs for performance, alignment, and governance.
• Expertise in Python and core modeling frameworks like Tensor Flow or PyTorch, Hugging Face, Pandas, and Num Py.
• Demonstrated experience leveraging AI‑assisted development tools (e.g., Anthropic Claude via GCP) to rapidly prototype experiments and accelerate research iteration.
• Proven ability to communicate complex mathematical and…