Pardhu Mattupalli

At Johnson & Johnson, I build clinical decision support and predictive analytics models for healthcare workflows, improving model response accuracy by 25% and accelerating clinical data analysis and decision-making.

I re-architected healthcare data pipelines in Python, PySpark, and SQL to cut processing time by 35%, and deployed deep learning, NLP, and transformer-based models for disease-risk prediction, patient outcomes, and clinical intelligence. I also engineer RAG pipelines and AI agents for context-aware clinical retrieval and automated document-processing workflows.

Previously at Infosys, I developed fraud detection, credit-risk scoring, and customer-intelligence solutions, improving fraud detection accuracy by 18%, reducing data processing time by 47%, and lowering prediction latency by 30%.

I bring 3+ years of experience delivering secure AI systems across healthcare and finance, from feature engineering and model evaluation through AWS deployment, MLOps automation, monitoring, explainability, and regulatory governance.