Software Development Intern

• Redesigned and implemented the ContactUs submission form for Prime Video’s studio partners, enabling efficient case submission through Slate and integration with the Paragon case management system • Implemented dynamics UI components using React and Java, automated partner type detection, and expanded routing logic, achieving a 98% reduction in misrouted tickets • Added structured custom fields (e.g., SKU, Region, Topic) in Paragon, enhancing support team triaging, filtering, and case visibility • Integrated a Generative AI assistant to provide real-time suggestions, improving self-service and reducing chatbot reliance • Streamlined backend data handling by integrating with Amazon’s internal Java-based case management system, optimizing schema alignment and data persistence via Apollo-deployed services

Software Development Intern

• Implemented filtering using Homography and Fundamental Matrices in OpenCV, improving robustness and accuracy by +30% • Integrated optimized Homography matrix computation directly into the pipeline, utilizing Eigen’s Levenberg-Marquardt algorithm • and confidence-weighted residuals for enhanced accuracy, resulting in a +40% increase in system accuracy • Automated workflows for data handling with TinyXML, enabling efficient reading, saving, and transformation of keypoints and matches. Developed visualization tools to dynamically plot matches, streamlining debugging and analysis • Conducted extensive testing with handmade ground truth cases, validating system performance across diverse scenarios and datasets • Identified algorithmic limitations, such as rotation invariance, and reviewed research papers to prototype solutions, laying the groundwork for further enhancements

Software Development Intern

• Created Python-based tools to track and analyze data flow and latency issues in base station ping packets. The tool has features to identify source and target IPs, and to extract PDCP IDs. Engineered comprehensive features for analyzing TTI data. • Designed algorithm to calculate frame and slot differences across various datasets, reduced analysis time from 5 hours to 15 minutes • Hands-on for over 1,000 test cases for various chip components including AALU, cCACHE, CXB, GFU, HPOE, HQM, ISA, OE, and Little core, enhancing testing protocols and efficiency using Python

Master of Science in Computer Engineering

Major courses: Cloud Computing, Natural Language Processing, Advanced Database Systems, Embedded Systems