HAT aims to bring together the best of what human tutoring and AI have to offer. Drawing on the profound benefits of a human tutor, the platform recommends challenging tasks, facilitates rich discussions, fosters relationships among students and tutors, provides feedback and guidance, and promotes collaborative learning.

Using learning engineering methods, the project aims to rapidly transform tutoring from a unimodal, one-on-one, human or technical solution to a multimodal, multi-party, human and computer synergy, reaching over 275,000 diverse, low-income students within 5 years.

Project Title

Hybrid Human-Agent Tutoring: Accelerating Middle School Math Achievement for Low Income Students

  Boulder, CO

Where it is used:

  • Chicago, IL
  • Greendot Public Schools, Tennessee
  • New York City, New York
  • Fairfax County, Virginia
  • Rhode Island
  • Washington, DC
  • Maryland
  • Charleston, South Carolina
Project Summary

What’s the problem that UC Boulder is trying to solve?

Frequent one-on-one tutoring can significantly help in reducing the differences in academic performance among students. But tutoring remains hard to scale. It’s also very expensive, and many tutoring programs rely on recruitment of tutors who may possess college degrees but lack formal training in education or licensure to teach.

What does the intervention do?

In combination with the proven Saga tutoring model, the program will automate tutoring to serve low-income and historically marginalized students. By working with teachers, the technology will be trained to offer differentiated, targeted feedback based on the type of mistake a learner has made. In doing so, learners are scaffolded to eventually explain their problem-solving processes, aware of where common errors can be made and avoided.

What is the wow factor?

To drive down the cost of tutoring and make it available to all students, the platform will use natural language processing to evaluate students’ self-explanations as they solve problems. For example, data captured in self-explanations can help identify that a student has a vocabulary gap or reading comprehension challenge. The intervention will also provide AI-driven feedback to students and tutors and have an equity analytics component to illuminate patterns related to inclusion and access.

How does it work?

The platform will build on the work done by Saga to build effective one-on-one tutoring. It aims to automate key student and tutor recommendations. It will focus on language recommendations by leveraging advances in natural language processing. It will also focus on creating improved feedback and scaffolding for students via AI.

Learn more at:

or contact Chris Yankee at chris.yankee@colorado.edu or Antonio Gutierrez at agutierrez@sagaeducation.org.

Team Members

Sidney D’Mello
Professor in the Institute of Cognitive Science and Department of Computer Science at the University of Colorado Boulder
Tammy Sumner
Director of the Institute of Cognitive Science and Professor in Computer and Cognitive Science at the University of Colorado
Jennifer Jacobs
Associate Research Professor at Institute of Cognitive Science at the University of Colorado Boulder
Peter Foltz
Research Professor at Institute of Cognitive Science at the University of Colorado Boulder