Paul Raff Email & Phone Number

Pittsburgh, Pa, Us

San Francisco, California, Us

Focused on the Search and A/B Experimentation areas of Data Science at Reddit - additionally providing thought leadership and technical direction throughout the Data Science and broader technical organization at Reddit.

San Francisco, California, Us

Bringing community and belonging to everyone in the world.

Hartford, Connecticut, Us

Tech lead and machine learning owner of the data and ML behind the next-generation Transformation Diabetes Care product, which optimizes diabetic care for hundreds of thousands of diabetics. Our novel approach identifies and plans a personalized approach to our members' next-best action as it relates to their diabetic care and well-being. Beyond diabetes, I provide technical leadership and strategy for the broader Analytics and Behavior Change organization, working with individuals and teams to achieve consistency and scale. Last but certainly not least, I'm an expert in healthcare data.

Redmond, Washington, Us

Program Manager Lead in the Core OS/Intelligent Edge Data Science and Data Engineering organization. I work daily with fantastic data scientists, data engineers, software developers and fellow program managers to utilize data to power progress for Microsoft's OS and Microsoft's cloud. I had six direct reports. Owned the data pipeline powering Windows 10 Monthly Active Devices, providing key data, analytics, and insight for senior leadership towards the push to 1 billion. Partnered closely with engineering and data science to standardize and make consistent the lifecycle of data science projects, greatly accelerating iteration cycles and loosening the engineering bottleneck. Inspiration: https://multithreaded.stitchfix.com/blog/2016/03/16/engineers-shouldnt-write-etl/Worked as part of a cross-organization effort to optimize the biannual Windows releases to serve to the right people at the right time. Managed a large-scale ML algorithm that scored hundreds of millions of devices daily and pinpointed issues related to the rollout. Owned the qualitative feedback pipeline, processing all feedback obtained through the Windows 10 Feedback Hub and other sources. Envisioned and drove a self-service framework for data scientists to plug in their own NLP pipeline without engineering bottlenecks.

Redmond, Washington, Us

Just as before, but now with a team!

Redmond, Washington, Us

Accelerating innovation through trustworthy experimentation and analysis.

Redmond, Washington, Us

Partnering closely with the Windows team to drive improvements to their Smart Search feature through data analysis and A/B experimentation. Improvements derived from the insights of myself and my colleagues have resulted in massive user satisfaction gains.Working on the general problem of understanding data *at scale*. As we move beyond Bing and to the rest of Microsoft with our data-driven culture and ability to extract insight from the data, it's important to be able to take a new data set and be able to automatically understand what the semantic meaning of the data set is with minimal human intervention. Implemented programs and systems that provide the ability to summarize and inspect new data sets.Leading design and implementation of experimentation summarization effort.Evangelizing the data-driven culture and constantly preaching the need to not only produce data, but the meaning within the data.

Redmond, Washington, Us

On the Bing Data Mining Analysis team, we work closely with all feature teams to help them plan, execute, and analyze high-fidelity, trustworthy A/B experiments on bing.com. Additionally, we work on the experimentation and metrics platform to create, analyze, and assess new metrics and methodologies in estimation.

Seattle, Wa, Us

Developing practical solutions for an immense supply chain. Trading off exact solutions to approximate models for approximate solutions to exact models. Giving expert, precise analysis of issues facing the largest retail supply chain on a daily basis.Was the driving force behind multiple fulfillment optimization solutions, enabling millions of dollars of savings in an already heavily-optimized supply chain setting. Turned real problems into abstract ones, solved the abstract ones, and then applied the abstract solutions into real solutions.Specifically, created the prototype and theoretical solution for real-time global optimization of order fulfillment, which led the move away from greedy fulfillment and realized cost savings of $80MM/year. Collaborated with other applied researcher colleagues and software developers from the whole supply chain stack to bring to production.

New Brunswick, Nj, Us

I was in the Department of Mathematics at Rutgers University under the advisorship of Professor Doron Zeilberger. Up until 2009 I was a Teaching Assistant, and afterwards I was a Graduate Assistant working on the Monitoring Message Streams project sponsored by IARPA.

While a student in the Department of Mathematics I taught a variety of courses, from Pre-Calculus to Combinatorics.

Princeton, Nj, Us

I participated in an intensive summer program to learn the ins and outs of math assessment.

I was a summer intern in the Shielding group, working on a project to overhaul their simulation software.

Ms, Mathematics

Ph.D., Mathematics

Bs, Mathematics

Bs, Computer Science

High School Degree