At Growfinix Technology, I architect and optimize machine learning preprocessing pipelines for patient records, engineer clinical features, and evaluate health risk prediction models. My Logistic Regression and SVM models achieved 88%+ predictive accuracy.
I also designed an Agentic RAG pipeline using LangGraph and the ReAct framework, with hybrid search and re-ranking that improved context retrieval precision by 25%. In a property price prediction project, I engineered listing features, explored geographic and non-linear drivers, and built models for web API deployment.
