Rasi Elluri

Ai Ml Engineer @UnitedHealth Group

US

Designed, implemented, and deployed end-to-end machine learning pipelines using PySpark, Databricks, and AWS SageMaker, reducing training time by 35% while scaling efficiently to process 50M+ healthcare records.• Developed, tuned, and validated predictive models using XGBoost, TensorFlow, and scikit-learn for patient risk stratification, improving early intervention accuracy by 18% and reducing hospital readmissions by 12%.• Built, optimized, and automated ETL/ELT workflows using Apache Airflow and Azure Data Factory to ensure reliable, scalable, and efficient enterprise data processing across multiple healthcare data sources and environments.• Created and integrated NLP pipelines using Hugging Face Transformers to extract valuable clinical insights from unstructured text, increasing relevant healthcare data extraction by over 40%.• Deployed high-performance machine learning models through REST APIs using Docker and Flask, enabling real-time inference with sub-300ms latency for critical healthcare decision support systems.• Implemented feature engineering best practices, data version control with Git/GitHub, and MLOps workflows, accelerating experimentation and deployment cycles by 25% while ensuring reproducibility.• Designed and delivered interactive Power BI and Tableau dashboards to visualize model metrics and healthcare KPIs, improving executive decision-making speed and accuracy by 20%.