Lead Data Scientist for Restrictions and Pull Payments Risk

• Completed the integration of a new Machine Learning platform for Data Science, onboarding 2 production ML models

• Led two other Data Scientists to improve data observability, overhauling 10+ data tables and making our data ready for analytics with AI agents

• Automated metric reporting for leadership of the Platform team (SVP-level), collaborating with 15+ cross-functional stakeholders

• Owned all data efforts (metric definition, ML prediction, impact sizing) surrounding restrictions, yielding xx% drop in users with bad restrictions and yy% reduction in restriction duration

• Overhauled the technical interview for Data Scientists at Coinbase

Traveling the world - 15+ countries on 4 continents!

I investigated drivers of creator monetization, focusing on factors which explain observed disparities between members of protected groups. This project involved advanced statistical modeling using R, in an area without any pre-existing solutions. Ranking drivers required a mix of machine learning and understanding of the product. Due to the novelty of the project, I also developed a customized way of measuring impact

Key achievements:

• Multiple factors influence monetization, which required coordination across several teams to gain a holistic view of the subject. Modeling efforts reached > x% R-squared performance, a y pp increase over previous efforts.

• Gave a presentation to executives, including the CEO, to present results and gain approval for future projects

• Findings have uncovered areas for future work, for my team and cross-functionally for the ML Fairness team

• Early results showed x% improvement on some key metrics

I worked on an ads product aimed at Small and Medium Businesses in "Next Billion Users" markets (South America, South-East Asia, Africa...). The four main pillars of the role consists of metrics reporting (dashboards, presentations), experimentation (track the metrics, make recommendations), predictive modeling and exploratory data analysis (to improve the product). Main stakeholders include product managers and engineers, up to the director level.

Key achievements:

• Supported the product team through y times growth in both user base and revenue, and full launch in x countries. Insights generated through analyses were key in unlocking this growth

• Built an experimentation infrastructure which cut down analyst hours by 90%, enabling the team to run 5+ experiments a quarter

• Trained a churn prediction model with TensorFlow, whose implementation contributed to improving our main retention metric by x percentage points

• Identified an ongoing bug in the product which impacted up to x% of users; recommended a fix which reduced its prevalence by y%

• Participated in a committee which reviews all external claims that Google Ads is making, reviewing 30+ quarterly to ensure accurate and reliable information is shared with the outside world