I currently work at Motional, where I work on ML data annotation and discovery tools in the frontend using React and TypeScript. My strengths include cross-team communication and zero-to-finished project ownership.
Owned frontend for multiple projects from start to finish, including an Audio Annotation Tool, a DTB annotation tracking tool, Train Station, and Job Metrics, all written in React using TypeScript.
•
Redesigned Traffic Light Detector tool, improving Q/A efficiency by over 5x, and annotation efficiency by at least 2x. Proposed and implemented numerous features to improve accuracy of annotation and user ease of use.
•
Proposed and implemented a major revision to our Scene Hunting tool to allow simultaneous playback of multiple camera video feeds at once, all perfectly sync’d, improving Scene Hunting productivity between 2x-6x.
•
Made significant improvements to LogMiner, including numerous bug fixes, features, and performance/usability enhancements.
•
Voluntarily wrote documentation for LogMiner, ReadMes for many projects, our team's repo, and backend annotation API.
Designed and developed a web-based Nuclear Cardiology reporting app using Angular 2.
•
Increased speed of the QUAD's floating license server by a factor of 8. Significantly increased stability with bug fixes and stress-testing software.
•
Made significant contributions to an iPad version of our flagship product, including massive rendering performance improvements (over 20x), numerous bug fixes, re-architecture of badly written classes, and AWS cloud integration.
•
Created MegaTest, automated regression-testing software to perform validation that key metrics do not change beyond expected ranges over time.
Designed and developed Midnight Mansion and a user-friendly level builder, which attracted an active fan base who still play the game to this day, over 16 years later. Ported game to SDL and OpenGL.
•
Created Midnight Mansion 2: The Haunted Hills
•
Created Insectoid, a Galaga-inspired space shooter
Designed and Developed Java GUI front-end for MACOS, a tool for analyzing aberrations in telescopes. Exceeded expectations in both quality and quantity of work completed, with little oversight.