Healthcare data professional with experience building machine learning pipelines, transformer architectures, and AI agent systems — with a focus on equitable outcomes in clinical settings.
| Project | Description |
|---|---|
| Bias Mitigation in Healthcare with Fairlearn | Mitigating racial bias in glioblastoma tumor detection using XGBoost and Fairlearn postprocessing — improved minority recall from 0.82 to 0.91 |
| Custom Decoder Transformer | Decoder-only transformer trained on patient diagnosis sequences (ICD-10 codes) to predict future diagnoses — treating medical history like a language |
| AI Agents with LangGraph & MCP | Async AI agent built with LangGraph and Model Context Protocol (MCP) for dynamic tool use |
| Azure ML SDK Pipeline | End-to-end ML pipeline on Azure Machine Learning using Python SDK v2 — XGBoost regression for blood pressure prediction |
| NLP Sentiment Analysis of Horror Movies | Sentiment analysis and K-means clustering of horror movie reviews using NLTK and TF-IDF |