Amazon
-
Software Engineer Intern 3 mos
Engineered an automated, AI-powered triage system for our team’s on-call ticket queue. By connecting an event-driven AWS pipeline to a multi-agent AI system built with Amazon Bedrock, the tool automatically intercepts incoming tickets and instantly generates a root-cause analysis and mitigation plan. This replaced a manual 15–30 minute debugging process per ticket and cut our average resolution time from 21 days down to just 8 days.
-
Software Engineer Intern 4 mos
Built diagnostic tools on a cross-platform debugging framework using a C++ to JavaScript WebSocket pipeline to establish live server-to-tool connections, enabling developers to monitor and test the Kindle application across devices for layout rendering, user interactions, analytics, and overall application behavior. This tool gave engineers one central place to monitor the app live, streamlining the debugging process and cutting bug reproduction time by 95%.
-
Software Engineer Intern 4 mos
Enhanced an in-house C++ rendering system for Kindle books by building custom UI components, including modular border styling and keyboard events. Worked across layout, drawing, and event handling to ensure books rendered smoothly and looked perfectly consistent on any screen.



