Developed end-to-end predictive models using Python, Scikit-learn, and TensorFlow to forecast patient medication adherence from structured healthcare datasets (>500K records). Built data ingestion and transformation pipelines with PySpark and AWS Glue, automating ingestion from pharmacy, clinical, and claims sources while maintaining HIPAA compliance. Optimized hyperparameter tuning and model training via MLflow and GridSearchCV. Implemented SHAP and LIME for explainable AI to interpret diagnostic predictions for clinicians. Deployed models as scalable RESTful APIs via Flask and Docker, integrated into the analytics platform for near real-time insights. Integrated AWS SageMaker pipelines for retraining and monitoring, ensuring version control across ML environments. Collaborated with data governance to align operations with HIPAA/GDPR requirements and presented dashboards with Power BI/Tableau for leadership.

Research Assistant at University of Texas at Arlington

May 1, 2024 - May 1, 2024

Conducted sentiment analysis research using BiLSTM and Word2Vec, achieving 90.3% accuracy on 50K+ text samples. Built SQL/PL-SQL data pipelines to preprocess institutional datasets, reducing latency and improving training throughput. Performed data mining, clustering, and classification with Python, Scikit-learn, and Weka. Generated synthetic data to balance underrepresented datasets and improve generalization. Documented methodologies for publications and presented findings with Power BI/Matplotlib for non-technical audiences.

Machine Learning Engineer at Dell Technologies

August 1, 2022 - August 1, 2022

Architected predictive models to forecast hardware component failures using Scikit-learn and XGBoost. Ingested 10TB+ of telemetry data via ETL pipelines with PySpark and Airflow. Improved model accuracy through feature selection, tuning, and cross-validation. Built real-time inference APIs with Flask and Docker. Implemented TensorFlow Serving for scalable deployment and monitored drift. Evaluated explainability with SHAP and translated findings for engineers and product teams. Collaborated with DevOps to integrate ML pipelines into CI/CD and documented governance and validation processes.