Q10

Developed various and ML systems software solutions for various AI/ML applications at Meta.

• Maintained FBGEMM, a low-precision math kernel library used for high-performance quantized server-side inference.

• Maintained and developed FBGEMM_GPU, a collection of high-performance PyTorch GPU operators and specialized GPU primitives used for deep learning recommendation systems (DLRMs).

Worked on infrastructure solutions for scalable device certification at Netflix.

• Designed and implemented edge- and cloud-based solutions to enhance device management and test infrastructure, enabling seamless full-lifecycle integration of the Netflix app on consumer electronics (CE) devices for both internal developers and partner organizations.

• Spearheaded the migration of the existing CE device integration testing infrastructure onto the replacement platform, bringing cloud-scale automated certification testing for partner devices.

Worked on distributed systems infrastructure solutions to suppport CI/CD at scale across various engineering teams at Apple.

• Architected and developed a company-wide cloud-native platform and ecosystem for delivering on-demand and multi-device CI workflows that is API-compatible with AWS CodePipeline.

• Led the development of a large set of Scala commons libraries for supporting rapid test-driven development of over 40 microservices that are built to have fault tolerance, high availability, and code correctness by construction.

Developed various HPC software solutions for various research groups in the Berkeley Lab.

• Collaborated with the Nuclear Sciences Division on cloud-based data analysis for large multisensor datasets using FastQuery indexing.

• Integrated FastBit into KISTI’s Astronomy Data Analysis Portal to enhance attribute-based access for massive scientific datasets, in partnership with KISTI, Yonsei University, and the University of Michigan.

• Enhanced the Warp PIC simulation package for non-static 3D mesh support and developed simulations to optimize ion beam parameters for heavy-ion experiments with the Accelerator and Fusion Research Division.