Hi, I’m Sai Nikhilesh Reddy, an Associate Machine Learning Scientist at Wadhwani AI. I work on clinical machine learning and generative AI systems for healthcare, combining research methodologies with production engineering to build tools that serve real users.

Research & Engineering

My work focuses on two main areas. First, I lead the development of a ML framework for keratoplasty care in partnership with LV Prasad Eye Institute (LVPEI). My work focuses on three core predictive formulations designed to support surgical planning and shared decision-making: estimating post-operative visual acuity to manage patient expectations, stratifying the risk of visit non-adherence to optimize clinical resource allocation, and modeling time-to-graft-failure. By leveraging Large Language Models (LLMs) to extract ground truth from unstructured clinical notes, I build scalable clinical decision support tools that enhance both patient outcomes and operational efficiency

Second, I build and deploy large language model systems at scale. I co-architected HealthVaani, a multimodal RAG assistant designed for 2 million frontline healthcare workers in India. The system handles voice queries in English and Hindi, serves 650+ active users across 6 districts, and has processed 11,000+ medical queries in pilot deployments. I built the evaluation pipeline using LLM-as-a-Judge methods and implemented safety guardrails for responsible AI deployment with government health stakeholders.

Publications

My research on evaluating robustness in LLM-based medical chatbots was accepted at the HEAL workshop at CHI 2025. You can read the paper here.

Background

I hold a Bachelor’s degree in Computer Science (Artificial Intelligence) from Amrita School of Computing. My technical work spans deep learning, natural language processing, retrieval systems, and deploying ML models in clinical settings.

Outside of work, I follow space tech, experiment with cooking, and share thoughts on AI and tech on Twitter. Feel free to reach out on LinkedIn or Twitter if you’d like to discuss machine learning, healthcare AI, or any shared interests.