Kevin dela Rosa

About Kevin dela Rosa

I am a Machine learning engineering manager and tech lead on Snapchat\'s Perception (Scan) team with over a decade of industry experience across the entire stack. I currently focus on problems related to visual search, image retrieval / matching, multimodal content understanding, big data, and the various signals that power the Snapchat\'s smart camera technology. I also have a background in building distributed web services, and have contributed to both the iOS and Android Snapchat apps for various featues. I\'m generally interested in building compelling user experiences powered by AI and content understanding. Previously @ Amazon, and Carnegie Mellon (Language Technology Institite). Resume: https://bit.ly/35A7UMm

Experience

1. Machine Learning Engineer

   Snap Inc.

   Jun 2016 — Present · Seattle, WA, US

   Drove & implemented camera unlocking experience that enables AR content to be triggered by scanning plain images (markers); unlocks thousands of community submitted lenses & high profile integrations like Cheetos scannable commercial (Superbowl LIV), Alex Israel x Snapchat (Art Basel), Vogue x Snapchat (Cannes Lion) Launched visual classification via Scan Module in Snapchat’s Lens Studio which gives AR creators the ability to recognize 500+ object types across 4 categories (Objects, Places, Cars, Dogs) within a Snapchat lens Drove development of Snapchat’s first in-house optical character recognition service used by visual search products and spotlight content moderation services Drove development of reverse image search capabilities for Scan, and deployed supporting infrastructure including focused web crawling and large scale web archiving / data warehousing solutions. Authored billion+ scale embedding similarity search engine that powers visual search use cases (e.g. fashion, product, lenses) and content-based machine learning data collection (e.g. visually or audio similar media retrieval) used to collect candidate training data for 50+ use cases (e.g. fine grained facial expression detection, body pose for skeletal tracking, cannabis / drug content classification, pet segmentation, audio scene classification, etc.) Trained & deployed multiple computer vision models powering Scan including fine grained scene classifier, sky segmentation, hand detection, logo detection; also contributed to backend services supporting Scan Contributed to backend & iOS/Android apps for context cards, widgets, search, messaging, friending & web Managed a mix of core infrastructure and machine learning engineers, actively recruited & defined job roles, grew team from 0 to 9 engineers in 7 months; defined & owned technical roadmap for Perception content understanding

Education

• Carnegie Mellon University

  Master's Degree, Computer Science - Language Technologies

  2009 — 2011

• The University of Texas at Arlington

  Bachelor's Degree, Software Engineering & Physics

  2004 — 2009