Building the foundational machine learning infrastructure that powers large-scale fine-tuned models — enabling intelligent document understanding, advanced computer vision, and agentic orchestration at enterprise scale.

As a core Foundry engineer, I led the technical design for both a file system primitive that improved resource retrieval and browsing, and an LLM-powered search tool that enhanced discovery capabilities across the platform. I played a major role in building a versatile command palette, optimizing navigation and command execution throughout the platform. Additionally, I developed a search interface to streamline workflows and provide advanced search and discovery features.

At Aggero, I optimized job processors for scalable influencer data analysis and developed brand recognition computer vision and influencer aggregation algorithms. I also enhanced developer and build tooling, and led the prototype of an original transformer-based, one-shot logo detection on 3D objects in videos, exploring new frontiers in brand recognition technology.

As a core engineer, I made significant contributions to the open-source community, mentoring in the Google Summer of Code (GSoC) and Google Code-In programs, and was recognized as the youngest GSoC mentor in 2017. My work addressed key issues, enabling full functionality for DVB and Teletext subtitle extraction from multi-program MPEG-4 live streams, facilitating the subtitle extraction for UCLA’s NewsScape archive video library. Winning Google Code-In 2016 allowed me to introduce high school students to the program, fostering their interest in open-source development and promoting collaboration across diverse teams.