Sagar Mandiya

Data Scientist @UnitedHealth Group

Delivered supervised ML models using AWS SageMaker, Lambda, and S3 to improve risk stratification and support clinical decision-making across healthcare systems.​ Developed a RAG-based clinical support engine using LangChain and OpenSearch, reducing physician response time by 22% in complex case workflows.​ Designed and optimized RESTful APIs for real-time inference, addressing latency issues in production ML services.​ Implemented caching and query tuning strategies that reduced API response time by 15%, improving application responsiveness and scalability.​ Applied causal inference and A/B testing to evaluate the effectiveness of healthcare policies, enabling better resource allocation.​ Generated interpretable insights that directly supported targeted interventions and public health strategy planning.​ Used LLMs to generate plain-language narratives from model outputs, improving adoption of risk insights by non-technical stakeholders by 40%.​ Ensured all solutions followed AWS Well-Architected Framework and met HIPAA and enterprise compliance standards.​ Aligned ML and RAG deliverables with business goals by clearly communicating results and trade-offs to clinicians, stakeholders, and leadership.