Shay Yazdani

Data Scientist @Optum

San Jose, CA, US

Developed a predictive model to forecast hospital readmissions using healthcare claim data, achieving 97% accuracy and 96% precision with Random Forest and Gradient Boosting models. Applied time-series cross-validation, PCA, and feature engineering to ensure robust performance and prevent data leakage. Explored GenAI techniques to enhance model interpretability with human-readable explanations for clinicians.Extracted actionable insights from large-scale healthcare data, uncovering trends and patterns to drive informed decision-making for improved patient outcomes- Developed and deployed an advanced analytical workbench, empowering clinical trial researchers with scalable, end-to-end data solutions- Partnered with cross-functional teams across healthcare and pharmaceuticals to design and implement data-driven solutions, bridging machine learning with real-world business impact- Provided analytics and machine learning consulting services to dozens of Optum clients, translating complex data into tailored, business-impacting solutions- Prototyped ML-based recommender systems (Keras) to optimize specialist referrals and accelerate diagnoses, leveraging real-world healthcare claims data.

Georgia Institute of Technology

Master of Science - MS, Data Science and Analytics

University of Tehran

Bachelor's degree, Biology

University of South Carolina

Master's degree, Neuroscience