International Best Researcher Award
I’m a Staff Data Scientist at Glassdoor, where I design and lead initiatives that use machine learning and analytics to improve user engagement, personalization, and growth. My focus is on turning complex behavioral signals into predictive insights that help users discover the right jobs—and help the business drive value through smarter lifecycle strategies. One of my most impactful projects involved building a machine learning system to personalize job alert emails using behavioral segmentation, latent intent modeling, and reinforcement learning for optimal send times. This resulted in a 40% increase in apply-starts and helped re-engage thousands of dormant users with more relevant, timely content. I’ve also developed predictive models to detect user fatigue, enabling us to pivot from frequency-based outreach to intent-driven messaging. This shift reduced unsubscribe rates and increased long-term retention. A segmentation framework I built also uncovered a high-value “content-first” user segment that didn’t apply to jobs immediately but drove long-term platform engagement and reactivation—challenging conventional assumptions about user value. Outside of engagement, I’ve led efforts to detect early drops in SEO traffic using time-series anomaly detection and behavioral drift forecasting, helping prevent revenue-impacting traffic loss from silent frontend issues. Before joining Glassdoor, I worked in fintech, insurance, and e-commerce, building pricing models, churn predictors, and fraud detection systems. My approach blends technical rigor with cross-functional collaboration, ensuring that models lead to action, not just insight. I’m passionate about building data science systems that don’t just optimize metrics, but serve people—whether through better job discovery, reduced friction, or more equitable communication strategies.