Google
- 2012 – Present
- Staff Software Engineer
- New York, New York
Tech Lead/Manager (TLM), Developer AI Research
Research and build AI-based developer tools.
- Initiated development and research on Google's first deployed ReAct agent program repair tool
- Worked across teams and functions to deploy APR in production, with thousands of fixes submitted
- Built evaluation sets, infrastructure, and metrics for program repair
Tech Lead/Manager (TLM), Speech Contextual Modeling
Improve speech recognition accuracy across surfaces using non-speech context, e.g., contact names.
- Managed research group (4 directs) on contextual modeling for E2E ASR
- Led team improving contextually-relevant phrase recognition in E2E ASR models
- Shipped novel techniques for increasing recognition accuracy
- Presented and published speech recognition research, both internally and externally
- Led area-wide migrations to new, state-of-the-art speech recognition architectures
- Deployed work in flagship products (Pixel, Assistant, Gboard, Cloud, Meet)
Tech Lead, Search Serving
Create and extend APIs for accessing structured knowledge bases in search and other products.
- Proposed, designed, led implementation of structured knowledge API serving 30k+ QPS since 2015
- Led design and development of domain-specific language for structured search result pages
- Proposed and implemented structured knowledge panel editor with live preview
Software Engineer, Crawler
Optimize crawling and indexing with new approach to predicting which URLs lead to identical content.
- Expanded initial proof-of-concept system to feature parity with previous, decade-old system
- Sped up system 5x (60 hours to 12) and reduced memory footprint 3x (36GB to 12GB)
Ph.D., Computer Science
University of California, San Diego
A novel approach to software verification through refinement type inference for C and OCaml.
2012
M.S., Computer Science
University of California, San Diego
2009
B.S., Computer Science
Pennsylvania State University
2006