SaiSindhu Beeravolu

About

Healthcare ML/AI engineer with production experience at UnitedHealth Group, including an end-to-end readmission prediction system built on 50M+ patient records that improved accuracy by 18% and reduced preventable readmissions by 12%. Also shipped a clinically grounded LLM/RAG referral generator with human-in-the-loop safety controls, showing strong depth in regulated, high-stakes AI systems.

Key Strengths

• Owned end-to-end healthcare ML system from experimentation to deployment and monitoring
• Built readmission prediction on 50M+ patient records
• Improved model accuracy by 18% using XGBoost and feature engineering
• Drove 12% reduction in preventable hospital readmissions
• Strong production ML monitoring with drift detection, calibration, and retraining
• Applied NLP and clinical notes to improve model performance
• Shipped safety-conscious LLM/RAG workflow for clinical referral generation
• Balanced latency, cost, quality, and safety in production GenAI systems
• Improved RAG reliability through retrieval, chunking, and prompt refinements
• Built reusable Python libraries that improved pipeline reliability and team velocity

Experience

Education

Skills

Languages