Using AI to Turn Grading into Learning

Gradescope Blog
Published in
3 min readJun 12, 2018


We are excited to announce that we’ve raised $2.75M for Gradescope! The round is led by Reach Capital, with GSV AcceleraTE and Ironfire Ventures joining our existing investors K9 Ventures, Freestyle Capital, and Bloomberg Beta.

We started Gradescope when we were PhD students at UC Berkeley. Even though we were teaching the futuristic subject of Artificial Intelligence, we were still grading exams the old-fashioned way: sitting around a table with big stacks of paper to pass around, red pens in hand, and boxes of pizza for fuel. It was logistically challenging, prone to inconsistent grading, and slow. Worst of all, we didn’t gain much insight into what the students actually learned — and the students often didn’t find our marks on their exam particularly enlightening, either.

There had to be a better way.

We built Gradescope to let instructors grade paper-based exams online. Gradescope lets instructors grade from anywhere, with consistent rubrics, and with AI assistance for grouping similar students answers together. In addition to exams, instructors can grade homework, bubble sheets, and even programming projects on the same platform. Most importantly, students get their work back much more quickly and with more actionable feedback. Furthermore, instructors receive helpful statistics as to what their students actually learned — we like to say that this turns grading into learning.

In the last year, we’ve hosted over 500,000 enrollments at over 600 universities, colleges, and high schools. All told, 40,000,000 students answers have been graded on Gradescope, with more and more of them first being grouped with AI-assisted grading. AI-assisted grading, along with other exclusive features, is now accessible to all instructors at over 20 institutions that have signed up for licenses in the last academic year. We’re honored to work with these schools, which span from small community colleges to some of the largest universities, and look forward to many more partnerships to come!

We’ve also had several more exciting things happen over the last year:

  • Our CEO Arjun was featured in the Forbes 30 and 30 and Cafe 100 lists.
  • We received a Small Business Innovation Research grant from the National Science Foundation to develop an AI system for grading 10x faster using deep learning. We have already submitted a report on our work to an academic conference, and look forward to deploying the innovation we developed in the product.
  • We presented a paper describing the fundamental Gradescope system at the Learning at Scale conference in 2017, and will be presenting a new paper on how higher-ed STEM instructors format their exams at Learning at Scale 2018 this June.
  • We grew from 8 people to 13 people.

We’re looking forward to the road ahead! If you’d like to join us, we are currently hiring for multiple positions: