Uday Kumar Swamy

Designed risk prediction models using Scikit-learn and XGBoost to improve patient-level forecasts and reduce false positives in internal healthcare review systems. Built NLP pipelines with Hugging Face Transformers and LangChain to extract unstructured clinical data. Applied SHAP and LIME to improve explainability and support HIPAA/compliance. Deployed models via FastAPI and Docker, reducing prediction latency and integrating into cloud-based patient reporting tools. Built ML pipelines with Airflow and AWS SageMaker Pipelines to boost throughput for large-scale healthcare datasets. Used MLflow for experiment tracking and lifecycle management across cross-functional teams and regulatory reviews.

Designed and implemented risk prediction models and NLP pipelines to improve healthcare forecasting and data extraction accuracy. Applied explainability techniques for compliance and trust, deployed models with FastAPI and Docker to reduce latency, and built ML pipelines with Airflow and AWS SageMaker Pipelines to enhance throughput. Utilized MLflow for model lifecycle management supporting cross-functional research and regulatory reviews.

Led design and development of ML solutions across 5 cross-functional teams in retail and financial services. Developed and deployed predictive models using Scikit-learn and PyTorch, achieving measurable improvements in churn reduction and customer engagement. Conducted market basket analysis and built a spam detection classifier to optimize marketing and communications. Implemented end-to-end CI/CD pipelines with Jenkins and AWS, and ensured GDPR-compliant data usage with drift detection and monitoring. Delivered scalable RESTful APIs for model inference with latency under 100ms and contributed to overall model reliability and fairness.