MSAI Graduated Student at University of Georgia Isha Patel is an accomplished AI/ML Engineer and Software Developer based in Georgia, United States, with over four years of experience in generative AI, data analytics, and business intelligence. Holding a Master’s degree in Artificial Intelligence from the University of Georgia and an MS in Information Technology from Uka Tarsadia University, Isha specializes in Retrieval-Augmented Generation (RAG), Large Language Models (LLMs), and Natural Language Processing (NLP). Her technical expertise spans Python, TensorFlow, PyTorch, and AWS, complemented by proficiency in visualization tools like Tableau, Power BI, and AWS QuickSight. Isha has made significant contributions across various roles, including her work at Model Lab in Athens, Georgia, where she enhanced transformer-based models like GPT and RoBERTa, improving system accuracy and efficiency. At Starbyte Infotech and Escrow Infotech in India, she developed scalable AI solutions, optimized ML pipelines, and deployed APIs for mobile and web applications, achieving notable improvements in prediction accuracy and processing times. Her projects, such as text summarization with AI agents and business intelligence analysis using Power BI, showcase her ability to tackle complex challenges with innovative solutions. A published author in the National Journal of System and Information Technology for her work on optical character recognition, Isha combines technical rigor with strong problem-solving and communication skills. Passionate about leveraging AI to drive impactful outcomes, she thrives in collaborative environments and continues to push the boundaries of scalable, data-driven technologies. Education: Uka Tarsadia University Master of Science - MS (Information Technology) 2014 - 2019 The University of Georgia Master of Science - MS (Artificial Intelligence) 2022 - 2024 Research Research Interests: LLMs, Prompt Engineering Degree Completion Date: Fri, 12/13/2024 - 12:00pm Dissertation/Thesis Title: Transformer-Based Embedding Models for System Similarity: A Comparative Study of RoBERTa and GPT - https://openscholar.uga.edu/record/26650?ln=en&v=pdf