Apr 2024 – Present · 2 yrs

• My organization covers applied science and research (causal inference, AI evals and explainability, econometrics), strategic deep dives/analysis, analytics engineering, distributed data systems, and visualization engineering for the Netflix member experience, personalization AI/ML, and content merchandising domains. We leverage and build AI tools across the stack for natural language access to analytics, knowledge management, and scaling user research through AI user personas. We have a deep appreciation for product thinking, and balancing customer needs with business objectives.

Dec 2021 – Apr 2024 · 2 yrs 4 mos

• My teams cover the following business domains: Payments, Partnerships, Messaging, Identity, Regional Growth, and On-Product Content Merchandising. Our functional expertise includes experimentation, observational causal inference, machine learning, and analytics.
• We partner with product and business area owners to identify and size opportunities, and evaluate the causal impact of projects and initiatives through the design, execution and analysis of randomized experiments/quasi-experiments, observational studies, and analytical deep dives. We are active thought partners to the business on strategy, roadmaps, OKRs, and performance evaluations. We develop, deploy and monitor machine learning models in production, and work with our engineering and platform teams to improve paved path solutions for machine learning practitioners.
• Payments work examples:
• ML models for fraud detection, payments charge scheduling, transaction routing, etc.
• Add and remove payment methods by country using A/B tests
• Analytical tools and deep dives to develop a card scheme tokenization system
• Partnerships work examples:
• Observational causal studies in partner settings where we can randomize users. Eg. value of a remote control Netflix button, new Fire TV UI redesign in 2021, firmware upgrade rollouts by Roku, etc.
• A/B tests to decide on the best UI placements and slots for Netflix, content personalization outside of the Netflix app, optimal signup and payments flows that begin outside the app, etc.
• Privacy-forward experimentation framework that allows Netflix to design and execute A/B tests on partner devices without needing to exchange individual-level data
• R&D examples:
• Sherlock: causal ML R package for segment discovery and analysis in A/B tests
• Novel effect estimators for quasi-experiments using robust inverse propensity weighting
• Fact stores for opinionated feature engineering, enabling real time inferencing, and eliminating skew between training and testing sets

Feb 2018 – Dec 2021 · 3 yrs 10 mos

• I lead a team of data scientists and analytics engineers to support the Payments and Partnerships verticals globally. We support product and business decisions through experimentation, modeling/ML, analytics, and reporting.

Jun 2016 – Feb 2018 · 1 yr 8 mos

• Growth experimentation and modeling for acquisition of new members internationally
• Lead data scientist on product partnerships, global payments, and integrated payments
• Optimization of signup and onboarding flows for new-to-Netflix and rejoining members
• Revenue and LTV maximization initiatives, free trial grant/deny rules, fraud and abuse checks
• Unsupervised learning and segmentation of markets