Artificial
Intelligence
Engineer.

I'm Guradesh Dhillon, a final-year Computer Engineering student at SVPCET, Nagpur, with a 9.3 CGPA. I learn by shipping, not just studying, and most of what's here is proof of that. My flagship project, CreditWise AI, is a full-stack loan approval platform: an XGBoost model tuned with Optuna, served through FastAPI, containerized with Docker, and tracked with MLflow. It pairs SHAP-based explainability with the Claude API to turn every prediction into a plain-English decision summary, alongside dashboards for what-if analysis and fairness auditing. On the computer vision side, I'm building DentalSeg AI, a two-stage tooth instance-segmentation pipeline using YOLOv8-Seg for detection, SAM2 for pseudo-labeling, and a rule-based FDI numbering engine for clinical labeling, grown out of a HealthTech internship where I ran the annotation pipeline from raw intraoral photos to COCO-format datasets. I'm also currently building Protocol-Alpha, an on-device AI privacy middleware for Android that runs Phi-4-Mini locally, so sensitive data never has to leave the device. Alongside these, I'm completing a Data Analytics internship under AICTE and Vodafone Idea Foundation's VOIS for Tech program, working through HR, agriculture, and operations datasets with Python, and I hold certifications including Oracle's Foundations Associate in AI & Agentic AI and Harvard's CS50 AI with Python. Outside of models and notebooks, I lead technical infrastructure as Event Technologist Head for the ACM CSE Student Chapter, because good engineering is as much about systems thinking as it is about code.