Overview
In 2026, I made the decision to jump from Netflix to project-based advisory work. I love to move fast and deliver measurable impact. My key strength is quickly identifying high-leverage opportunities and designing simple yet valuable 0-to-1 solutions. I am so lucky that I now get to craft my career around partnering with teams on this type of problem solving. Below are just a few examples of how I’ve done this in the past. Please reach out if you’d like to discuss a challenge your team is facing!
Project Examples
At Netflix, I recognized redundant query work and built an AI-powered app that enabled teams across the business to build custom audiences. Within a few months of launch, the app was powering billions of home screen impressions and messages per month and was being used by over 80 internal users per month (and over 30 per week).
At Glassdoor, I led the home feed ML recsys team and saw an opportunity to better serve our new users. With my team, I built a simple cold start algorithm that boosted new engaged users by 35% within a few weeks.
At Concentric (now SciOS.ai), I recognized the new user pain in onboarding to our product. I streamlined the experience and underlying model calibration, removing 85 clicks from the model onboarding process and cutting time-to-calibrated-model from weeks to minutes.
In this new phase of project-based work, my first few projects center on improving perinatal care in the US. Drawing from my experience as a Head of Product & Chief Data Scientist, I’m building a suite of models to help a perinatal mental health company scale. Leveraging my causal inference and research skills, I’m diving into ferritin testing in pregnant and postpartum women to advocate for improved testing in a country that has (IMO) over-normalized fatigue, brain fog, and low mood among new moms.
About Me
I spent the past 10 years in tech, holding a range of IC and leadership roles. At Netflix, I was a data scientist focused on scaling the Messaging experimentation program through standardization, automation, and self-serve experimentation platform design. At Glassdoor, I was the Senior Director of Machine Learning & Applied Science, where I led the recommendation system as well as the internal growth model to understand high leverage product investment opportunities. Earlier at Glassdoor, I was the Director of B2B Data Science & Engineering, where I worked on a range of sales claims models, sales data pipelines, and initiated the overhaul of Glassdoor’s product analytics suite from “flying blind” to industry-recognized. Prior to Glassdoor, I was the Chief Data Scientist & Head of Product at Concentric (now SciOS.ai), where I loved seeing the tangible value of merging Data Science & Product by streamlining our UI and models to improve the user experience while also improving model fidelity.
I have a PhD in Public Policy from Harvard, where my research applied econometrics, statistics, and machine learning to data from Internet platforms. My dissertation Algorithms and Applied Econometrics in the Digital Economy included algorithmic work on nowcasting and dynamic pricing as well as applied econometric research on the value of financial assistance in online learning.
I am passionate about making quantitative methods feel intuitive and accessible to people of all backgrounds. This shows up in my daily work, but also in more formal teaching engagements. For example, I was an Adjunct Lecturer at Harvard, teaching the Masters in Public Policy first year course on data, econometrics, and machine learning (API 202: Empirical Methods II).
With this diversity of experience, I have deep cross-functional empathy and a broad toolkit helps me find an efficient solution to the problem at hand.
If you would like to connect about a potential engagement, please reach out. At this time, I am accepting new clients for early 2027.