Jon Noronha

Jon Noronha

About

Detail

Co-Founder
San Francisco, California, United States

Timeline

work
Job
auto_stories
Publication

Résumé

Jobs verified_user 0% verified
  • Gamma
    Co-Founder
    Gamma
    Nov 2020 - Current (5 years 8 months)
    We're inventing a new medium for business communication. Write like a doc, present like a deck, discuss live or async. Learn more at https://www.gamma.app/
  • Gamma
    Co-Founder & Chief Product Officer
    Gamma
    Nov 2020 - Current (5 years 8 months)
    We're inventing a new medium for business communication. Write like a doc, present like a deck, discuss live or async. Learn more at https://www.gamma.app/
  • Optimizely
    VP Product
    Optimizely
    Jan 2020 - Nov 2020 (11 months)
  • Optimizely
    Senior Director, Product Management
    Optimizely
    Dec 2018 - Jan 2020 (1 year 2 months)
    I help companies like IBM, the Gap, HP, and The New York Times build a culture of constant testing and learning, powered by the world's leading experimentation platform.
  • Optimizely
    Director, Product Management
    Optimizely
    Oct 2016 - Dec 2018 (2 years 3 months)
  • Optimizely
    Senior Product Manager
    Optimizely
    Jun 2015 - Sep 2016 (1 year 4 months)
  • Optimizely
    Product Manager
    Optimizely
    Jan 2014 - Jun 2015 (1 year 6 months)
  • Microsoft
    PM, Image Search
    Microsoft
    Aug 2011 - Dec 2013 (2 years 5 months)
    Product manager for Bing's Image Search, coordinating engineering teams across Seattle and Beijing to rethink visual search. We developed cutting-edge technology in machine learning, distributed systems, and image processing and combined it with great design based on usability studies, constant A/B testing, and quantitive analysis. My job was bringing it all together to ship new features to shake up the search market.
  • P
    Lead Author
    PlateMate
    Aug 2010 - May 2011 (10 months)
    Invented a new system for getting high-quality data from crowd workers on Mechanical Turk and used it to build an app that lets you take a photo of your lunch and see you how many calories you ate. We built a Python/Django framework for breaking up complex tasks into simpler problems and designed a sequence of interfaces for workers to identify foods in a picture, match them with a nutrition database, and measure portions. We ran several studies to prove our system was comparable to experts in accuracy, easier to use than traditional food diaries, and less error-prone than amateur calorie counting. Published in UIST 2011 and cited in 180+ papers so far.
  • Microsoft
    PM Intern, Bing Relevance
    Microsoft
    May 2010 - Aug 2010 (4 months)
    Devised a new system for obtaining manual judgments of images and videos for competitive analysis and ranker training data. Designed crowdsourcing workflows, quality control mechanisms, and task interfaces to maintain data quality while cutting costs by 60% using Amazon Mechanical Turk, saving $1 million per year.
  • Harvard University
    Teaching Fellow, Computer Science 50
    Harvard University
    Sep 2009 - Dec 2009 (4 months)
    Taught 15 students the basics of programming and computer science, focusing on C and PHP and concepts like recursion, data structures and algorithms, time complexity, and pointers. Taught section, held office hours, and graded problem sets. Earned a Certificate of Distinction in Teaching based on great reviews from students.
Education verified_user 0% verified
  • Harvard University
    Bachelor's degree, Computer Science
    Harvard University
    Coursework in machine learning, artificial intelligence, HCI, robotics, auction theory, databases, algorithms, theory of computation, and systems. Published research in crowdsourcing and human-computer interaction. Additional courses in economics, history, philosophy, and literature.
Publications verified_user 0% verified
  • P
    Platemate: crowdsourcing nutritional analysis from food photographs
    Proceedings of the th annual ACM symposium on User interface software and technology UIST
    Oct 2011
    What if you could take a picture of your lunch, and have your phone tell you how many calories were on your plate? We built a crowdsourcing system to do just that, using workers on Amazon Mechanical Turk to label foods and estimate portions. Prior work in crowdsourcing mostly focused on simple work on toy problems -- we showed how to build a complex workflow with quality control to solve a real problem in the world. Based on a week-long study, our system was easier to use than self-reports and nearly as accurate as trained dietitian.
This is a community-created genome.