Group Tech Lead on Instagram, Relevance and Machine Learning Systems

• Proposed and created Product Mixer (open source http://bit.ly/productmixer), Twitter’s next generation product and recommendations real-time serving platform. Co-founded team, led platform development, and scaled adoption to 150+ product use cases replacing 10+ bespoke product platforms.• Led modernization and large-scale rewrites of Home Timeline (open source https://bit.ly/homemixer), Search, Social Graph, User Profile, Notifications, and Explore onto the Product Mixer platform, which resulted in a multiples increase in development velocity, -__% latency/TTFB, -__% capacity usage, +__% ad impressions, and +$___mm revenue.• Led design and development of TikTok-style immersive video viewer and recommendations. The system powers the Twitter video viewer and Explore see blog https://bit.ly/twittervideoviewer

• Co-architected and built a scalable, unsupervised machine learning system to consume the Twitter Firehose (~6k Tweets/s, 50k+ entities/s) and detect events in real-time. Co-founded team and scaled adoption to powers Twitter Trends, Events, and Explore products globally.• Published at KDD 2019 “Real-Time Event Detection on Social Data Streams.” https://bit.ly/eventpaper and granted U.S. Patent https://bit.ly/eventpatent• Co-created hackweek project which became the first example of real-time, algorithmic curation of Tweet Timeline Event summaries at Twitter. Co-founded team and scaled the product to power +5000 NFL, NBA, MLB, Soccer sports game Twitter Events annually.

• Led effort to stabilize and optimize Twitter Moments/Explore (~20 services), resulting in -__% latency, -__% capacity usage, and company-worst incident frequency to zero.

Led software engineering organization delivering safety-critical command & control and VoIP telecom products in the public safety, air traffic control, and defense sectors.