Senior Research Scientist - Personalization

Yahoo Mail is the ultimate consumer inbox with hundreds of millions of users. It’s the best way to access your email and stay organized from a computer, phone or tablet. With its beautiful design and lightning fast speed, Yahoo Mail makes reading, organizing, and sending emails easier than ever.

A Little About Us

The Mail Intelligence platform is building the next generation of services enabling Yahoo to deliver deeply personalized content to hundreds of millions of users. We process several petabytes of data, leveraging cutting-edge algorithms to extract knowledge and interconnect information from billions of messages. Our work spans NLP, deep learning, and massive-scale data processing. You will join a high-impact team of engineers and designers transforming the "ultimate consumer inbox" into an AI-native personal assistant.

A Lot About You

• Expertise in developing and applying state-of-the-art machine learning (ML) and deep learning (DL) models to solve massive-scale architectural challenges.

• Advanced ability to transform theoretical research into production-grade applications, specifically in personalization, ranking, and Generative AI.

• Commitment to pushing the boundaries of innovation while maintaining rigorous standards for model performance and data privacy.

• A mindset focused on automation; you view AI as a tool to accelerate your own research and optimize system-wide efficiencies.

• Collaborative leader capable of influencing technical direction across multiple squads and mentoring junior researchers.

Responsibilities

• Define the research strategy and lead research for innovative algorithms in personalization, ranking, and information retrieval for 100M+ active inboxes.

• Build AI-augmented experimentation pipelines that utilize automated tracking (e.g., Weights & Biases) to increase research velocity and model reproducibility.

• Design, fine-tune, and deploy state-of-the-art transformer and generative architectures, ensuring they are optimized for sub-second production latency.

• Lead the integration of AI-assisted development tools (e.g., GitHub Copilot) to streamline the transition from research prototype to production-ready code.

• Develop scalable, self-healing workflows for processing petabyte-scale datasets, replacing manual monitoring with AI-driven diagnostics.

• Establish protocols for the evaluation and validation of AI-generated outputs, focusing on detecting hallucinations and ensuring the safety of user-facing generative features.

• Stay at the forefront of the research community, driving Yahoo’s thought leadership through high-impact publications, patents, and conference presentations.

• Serve as a "prime strategist" for Mail Intelligence initiatives, mentoring junior scientists and aligning research goals with broader business objectives.

Basic Qualifications

• PhD in Computer Science, Statistics, Applied Mathematics, or a related field (Master's degree accepted with 7+ years of relevant experience).

• 5+ years of professional experience in machine learning or deep learning, with a focus on personalization or recommendation systems.

• Proficiency in fine-tuning Large Language Models and prompt engineering  to solve domain-specific tasks.

• Expert-level Python skills and hands-on experience with PyTorch, TensorFlow, and the Hugging Face ecosystem.

• Experience using AI-driven productivity tools to accelerate technical documentation, code refactoring, and research synthesis.

Preferred Experience

• Familiarity with Reinforcement Learning (RL) for ranking and recommendation optimization.

• Experience architecting solutions on cloud platforms like GCP (Vertex AI) or AWS (SageMaker).

• Proven track record of industry-level technical influence (e.g., top-tier conference publications like NeurIPS, ICML, KDD).

• Experience with large-scale distributed data processing (e.g., Spark, Beam).

 

The material job duties and responsibilities of this role include those listed above as well as adhering to Yahoo policies; exercising sound judgment; work