• Talks & publications
  • -->

    flyby

    flyby is an adaptive learning kit that can be dropped into any learning platform, LMS or application. The end result is your learners gets real-time, personalized recommendations on your content.

    Motivation

    In our Fly-by-Wire project, we developed an AI app for instructors and students. The app gives students formative assessment questions; it measures their performance on fine-grained learning outcomes, and intelligently responds to their mistakes by giving them more fundamental outcomes to master. flyby is the piece of the artificial intelligence that guides the student.

    How flyby works

    flyby requires two inputs: your collection of content and an underlying outcomes and content map. With these inputs, flyby computes the next outcome the learner should be navigated to, and serves up the content linked to that outcome. flyby runs client-side on the browser or server-side in Node.js.

    More documentation is coming soon in Summer 2017. Star our repo to be kept updated on code releases.

    Questions? Please email our lab's director, Prof. Karen Willcox [kwillcox@mit.edu].