I lead Python Language & Runtime org, consisting of 3 teams (25 members) of industry experts and managers in compilers, programming languages, type systems, and high-performance systems. We drive the next-generation, performance-optimized Cinder runtime for Python, as well as its integration into business-critical systems across our infrastructure. Pyre, our high-performance language services ecosystems, drives every Python developer's day-to-day experience, from type checking to privacy enforcement. We gatekeep the holistic end-to-end Python experience, designing whole-system overhauls for AI, build, deployment, codebase evolution, and more. I devise and lead the strategy to make Python the most productive language in use at Meta.

Led creation, development, and success of 3 teams (25 members) of industry-leading researchers, tech leads, and managers in interpretable and probabilistic modeling. Developed next-generation infrastructure to enable at-scale Bayesian learning, pioneered multi-year strategic initiatives around key research areas, and partnered with major company initiatives to drive end-to-end productionization of this cutting edge research for Ads, Experimentation, Integrity, Efficiency, and many more domains.

Full-stack engineer contributing to Google Voice team, Project Fi, and Hangouts. Tech led (pod of 6) an overhaul of the Android Google Voice experience. Overhauled backend calling services, performing a migration of 11 billion calling entities to improve calling efficiency and provide a modern platform to build innovative new services. Designed new services to enable seamless Wifi-to-cellular handoff for Project Fi calling. Enable Project Fi roaming call routing. Organized the Outbound Caller ID (Android Hangouts) and First Minute Free (cross-platform) features as the release engineer.

Paid research and development position with an agile development team under Harvard CTO Jim Waldo. Developed a large-scale analytics engine to provide analytics, inference, and visualizations for HarvardX, Harvard's division of edX.

Owned, optimized, and refactored the A/B Testing Framework on Android. Overhauled framework APIs to improve usability, provide more efficient and configurable logging, and introduce new features, including multi-account support and automatic restart when the app is sent to the background. Reduced Facebook’s overall mobile server load by 4% by batching A/B testing requests and logs. Reduced framework startup from 500ms to 200ms via lazy initialization and integration of a streaming storage representation.

Owned the Comment Replies feature, a major feature to promote high quality engagement on high-traffic content. Built a scalable backend service and storage to represent comments and threaded conversations, and performed a large-scale backfill to upgrade existing comments to support replies. Implemented a performant web frontend for comment replies powered by ReactJS. Integrated machine learning and computational linguistics to curate comments and replies for public content.

Paid teaching position for Harvard's most popular course, CS50. Led recitation for a class of 15 students. Conducted office hours with other teaching staff for the nearly 700 students enrolled in the class. Also responsible for grading assignments and promoting tech culture on campus.

Worked alongside software engineers in Shanda's Research and Development team to develop out an infrastructure to support big data-powered image classification of physical objects. Researched advanced computer vision techniques such as Scale-Invariant Feature Transform (SIFT) and Speeded Up Robust Features (SURF) to improve computer identification of Haar-like features. Integrated algorithms based on these approaches to develop a cascading feature detection architecture that exploited incremental learning over millions of images.

Research position under Dr. Charles Reinholtz in autonomous aerial robotics. Researched wireless protocols and optimizations under the Joint Architecture for Unmanned Systems (JAUS). Developed performant wireless communication protocols as part of an embedded system in a single-rotor autonomous aircraft using LabVIEW, which competed in the International Aerial Robotics Competition.

Independently developed a reactive web site to streamline interdepartmental communications, replace the email-based task management system, and automate common workflows. Worked with individual departments to incorporate this solution into employees' business operations.

Research associate under Dr. Elliot Douglas. Collaborated with graduate researchers to develop an enhanced fiber-reinforced polymer as a structural improvement agent for cementitious superstructures. Modified polymers to produce new strains of organic reinforcement agents, and conducted physical and computational tests to assess bond strength. Produced a research paper and presented lectures to graduate students.