I built a multimodal skin disease detection and diagnosis support system using YOLOv12s and symptom-based Logistic Regression. It achieved 79.17% diagnostic accuracy, and I deployed the model as a Dockerized FastAPI service on AWS.
For Cognifile AI, I built document intelligence features with RAG, Sentence Transformers, and Qdrant. I integrated the Gemini API to answer questions using retrieved document context.
I developed SupportIQ AI to classify support tickets, detect sentiment and urgency, and generate context-aware responses using Gemini and RAG. The platform includes LangGraph orchestration, FastAPI backend APIs, and a React dashboard.
At Liaquat University of Medical and Health Sciences (LUMHS), I implemented, configured, and tested the Student Information Management System (SIM), Moodle LMS, and TAO Assessment Platform. I also contributed to backend development and application integration at JAM Web Services, and gained AI and backend experience through internships at Evolvian Softwares (Pvt.) Ltd. and Mehran Sugar Mills Limited.

