Designing end-to-end ML systems for imaging, biosignals, and clinical data — from MRI-PET pipelines to EHR NLP models — with FDA-compliant workflows and measurable impact.
Building human-centered AI agents that handle complex processes, from HR compensation reviews to real-time biofeedback coaching. Skilled in prompt design, orchestration, and scalable deployment.
Developing robust models across domains: computer vision, NLP, transformers, and semi-supervised learning. Focused on turning raw data into clear, actionable insight.
Blending art, design, and technology through immersive installations, audio-visual systems, and experimental AI — creating experiences that spark curiosity and connection.
Experience spanning biomedicine, healthcare data, enterprise AI, and agent design.
Designed and deployed end-to-end ML pipelines for Parkinson’s stem cell trials, including MRI-PET analysis and Vision Transformer–based cell segmentation. Reduced manual processing time by 70% while delivering FDA-compliant outputs.
Created LLM-driven HRV biofeedback and CBT coaching agents integrating physiological sensing with NLP-based guidance. Increased user engagement by 10% and improved real-time interaction quality in mobile apps.
Built LLM-powered QA pipelines and accessibility ML systems, improving the AWS console for thousands of users. Cut customer support costs by 10% and standardized ML deployment practices across teams.
Created LLM-driven HRV biofeedback and CBT coaching agents integrating physiological sensing with NLP-based guidance. Increased user engagement by 10% and improved real-time interaction quality in mobile apps.