Leading an NSF-funded project ($150,000). First author of a research paper on automating CI troubleshooting using Bayesian group testing and Bayesian optimization (in preparation) for submission to cloud systems conference∗ Built BsoZebraConf, by augmenting the leading project for detecting heterogeneous-unsafe configurations with Apache Spark-based parallelized Bayesian test selection in Java and Bayesian optimization in Python∗ Reducing testing time by 61.8% with a 1% false positive and 0% false negative rate.∗ Utilized Docker for containerization and Bash for scripting. Managed builds and JUnit test with Maven.∗ Conducted distributed testing on AWS EC2 instance and a Cloudlab ClusterLead and manage a student team of 6 to compete in SC’24’s IndySCC Competition (Fully Funded)

• Engineered data processing portal utilizing NumPy and Scikit-learn to analyze frequency datasets. Achieved 93.3% reduction in manual processing time (15 min to 1 min)• Spearheaded frontend load-time optimization using TypeScript and Next.JS, reducing load time by 21% (2.8s to 2.2s)• Created REST APIs using TypeScript form email verification, group creation, role-based access control, etc.• Developed a managerial portal leveraging retrieval-augmented generation techniques with LlamaIndex to perform sophisticated comparative analysis. Significantly enhancing evaluation accuracy and efficiency