At Ventas Softech, I developed a loan prediction model for a banking client, comparing supervised algorithms and fine-tuning a Random Forest classifier that reached 89% prediction accuracy.
I also engineered a RAG pipeline for classical Sanskrit documentation, from text preprocessing and vector indexing in ChromaDB to retrieval tuning for scriptural references. I serve model predictions through FastAPI endpoints.
At Hisan Labs Pvt. Ltd., I worked as a Data Scientist Intern, handling daily data cleaning and validation workflows and developing supervised models with Python and TensorFlow. The diagnostic models’ reliability improved by 22%.
My projects include a multilingual RAG data analyst, a book recommendation engine, an MRI brain tumor detection system, and a healthcare analytics ETL pipeline. The brain tumor classifier achieved 99.08% accuracy, and I deployed it as an interactive web application.
