The CrowdGrader advantage

  • Simplicity: No complicated full-course setup. Just create an assignment, and add your students -- or share with them a URL where they can self-enroll. Read more.
  • Incentives: The grades given to students depend not only on the homework grade they receive, but also on the quality of their reviewing and grading work.  This is crucial to ensure the quality of the peer-review process.  Read more.
  • Secure and scalable: CrowdGrader is implemented on a dynamically scalable infrastructure, that can handle large class sizes and the activity spikes that occur as submission and review deadlines approach. All data is replicated in multiple geographical locations to protect against loss. Read more.
  • Easily compare submissions.  In crowdsourced grading, every grader sees only a small subset of the submissions.  To help in detecting plagiarism, CrowdGrader implements a powerful similarity checker for submissions.  Read more.


CrowdGrader is hosted on the Google Cloud for maximum reliability and availability. All data is geographically replicated, to ensure that no data is lost even in case of an infrastructure disaster. The infrastructure is fully and automatically scalable. In this way, if you have a very large class (thousands of students, or more), and everyone tries to submit as the submission deadline approaches, the computational resources of CrowdGrader will automatically scale up to absorb the usage spike.

Who we are

CrowdGrader is an offshoot of the research of professor Luca de Alfaro at UCSC.  CrowdGrader was originally created as a tool to experiment with crowdsourcing algorithms and incentives for collaboration.  After using it successfully in his classes on Android and Web development, professor de Alfaro started CrowdGrader LLC as a way to make the tool accessible by anyone, anywhere.