Yashveer User

I've shipped DocSense, a full-stack AI document assistant that lets users upload PDFs and receive page- and chunk-cited answers grounded in their documents. I built its FastAPI backend and REST endpoints, connecting a React frontend to ingestion, Chroma DB retrieval, and OpenAI-powered generation.

I've also built a serverless AWS data deduplication pipeline using Lambda, S3, and MD5 hashing to detect and remove duplicate files. This work was published as Data Duplication Removal Technology Using AWS Services in INT-JECSE.

My work spans machine learning and analytics as well: I trained an anomaly-detection model on roughly 25,000 synthetic physiological records, achieving 95% cross-validated accuracy, and created a PyQt5 dashboard for live visualization and alerts.

As a teaching assistant for Database Management Systems and Operating Systems, I taught SQL and database fundamentals, developed course materials, and supported students one-on-one. I'm completing a Master of Computing and Information Systems while building practical skills across Python, SQL, dbt, Snowflake, AWS, and data modeling.