Hi, I’m Mashiur, a junior at Vanderbilt University majoring in Computer Science and Mathematics. I love building impactful software that bridges AI with real-world needs. Recently I built an AWS Bedrock chatbot powered by Claude 3 Sonnet that analyzed 1,000+ SEC filings, improving Q&A accuracy by 60% through a context-aware pipeline. At the Vanderbilt School of Nursing, I developed a generative AI chatbot to streamline internal grant analysis, and I led workshops for faculty to adopt AI tools. I’m also working with Yrikka, an AI safety startup, on a Cornell Tech project to close the synthetic-to-real data gap by fine-tuning YOLO models on synthetic data for more robust object detection in autonomous systems. Through fellowships like Cornell Tech AI/ML, I’ve strengthened my expertise in scalable systems, applied machine learning, and full-stack development. I’m seeking opportunities to contribute to building reliable, high-impact applications that solve meaningful problems.