Hi, I'm Pooja š
Data Scientist | AI Enthusiast | Experimentation & Optimization Expert
Explore My Work
Data Scientist | AI Enthusiast | Experimentation & Optimization Expert
Explore My Work
Iām a passionate data scientist specializing in machine learning, causal inference, and AI-driven experimentation. I love transforming complex datasets into actionable insights and elegant visualizations.
š
May 2025 ā Aug 2025
š San Francisco, CA
Developed supply-demand forecasting models in Python/SQL/Excel on large-scale transactional datasets, improving rider-driver matching and reducing simulated wait times by 15%. Designed and executed pricing & incentive experiments using A/B testing + CUPED, generating statistically valid insights and achieving a rider conversion lift of +8%. Prototyped optimization algorithms (LP/MIP with OR-Tools) for driver allocation, boosting fulfillment rates during demand spikes by 12% and improving marketplace efficiency.
š
Feb 2024 ā Jul 2024
š Hyderabad, India
Built automated SQL-to-BI ETL pipelines supporting recommendation engines, raising CTR by 12% through more accurate targeting and personalization. Applied advanced causal inference methods (Diff-in-Diff, Propensity Score Matching) to evaluate incentive effectiveness, producing actionable insights for regional rollouts. Implemented ML model monitoring with drift detection, ensuring long-term stability, accuracy, and reliability of production systems.
Built scalable churn prediction pipelines using PySpark and XGBoost, achieving 92% AUC.
Audited ML fairness using SHAP & Fairlearn, ensuring ethical model deployment.
Forecasted flight ETAs using ARIMA, Prophet, and LSTM for better operational planning.