-Co-authored and submitted to AMCIS, research paper, “Unveiling Language Mastery: Exploring LargeLanguage Models for Academic Research”, pioneering a two-step generative clustering approach usingLLMs to introduce an advanced topic extractor for academic research. -Evaluated and enhanced the performance of K-means clustering and LDA algorithms, achieving a 14% improvement in clustering quality measured by key metrics consisting of Silhouette Score, Davies-Bouldin Index (DBI)… Show more -Co-authored and submitted to AMCIS, research paper, “Unveiling Language Mastery: Exploring LargeLanguage Models for Academic Research”, pioneering a two-step generative clustering approach usingLLMs to introduce an advanced topic extractor for academic research. -Evaluated and enhanced the performance of K-means clustering and LDA algorithms, achieving a 14% improvement in clustering quality measured by key metrics consisting of Silhouette Score, Davies-Bouldin Index (DBI), Calinski-Harabasz Index (CHI), and Dunn Index.-Conducted comprehensive first and second-round interviews to evaluate prospective candidates for the AI4Business Lab, accessing their proficiency in machine learning, coding, and essential soft skills. Show less

-Implemented Mixout regularization techniques during the fine-tuning phase of LLMs, injectingcontrolled randomness to prevent overfitting, improve the model’s ability to generalize from diverseinputs, and enhancing the model’s ability to capture data relationships.-Developed Python script to extract titles and bodies as key value pairs from dataset of 100+ PDFs, transforming the data into an organized JSON file, scoring key terms using TF-IDF and identifying most common n-grams… Show more -Implemented Mixout regularization techniques during the fine-tuning phase of LLMs, injectingcontrolled randomness to prevent overfitting, improve the model’s ability to generalize from diverseinputs, and enhancing the model’s ability to capture data relationships.-Developed Python script to extract titles and bodies as key value pairs from dataset of 100+ PDFs, transforming the data into an organized JSON file, scoring key terms using TF-IDF and identifying most common n-grams to enhance data analysis.-Conducted an extensive literature review on LLMs and their integration with deep learning methodologies. Show less

-Reduced end-to-end test case job run time by 78% in CI/CD pipeline by utilizing parallel execution using Python and YAML.-Enhanced EC2 instance compliance by attaching GP3 volumes, resizing instances, and attaching required security groups,improving security and cost efficiency by 20%.-Increased team productivity by leveraging LLMs and RAG to generate story and task guidance in new codebases

-Implemented and delivered engaging programming, robotics, circuits, and mechanical systems workshops for diverse groups of students of 30+.-Instructed students on the fundamentals of programming, covering essential topics including loops, recursion, and data structures.-Provided individualized support to students, offering guidance and feedback on their coding, engineering, and science projects and assignments.