Erik Harpstead

Erik Harpstead is an associate research professor at Carnegie Mellon University’s Human-Computer Interaction Institute whose work uses games to study learning and human-AI collaboration. He completed his doctorate at CMU in 2017, where his research used game replays and an AI player model to examine educational game design. He helped develop the Apprentice Learner Architecture and coauthored a 2020 method that let teachers build AI tutors through examples and corrections. Through Dice Adventure, his team created a setting where people and AI agents had to coordinate using their different information and abilities. He served as general chair of ACM CHI PLAY in 2025.

Now at: Carnegie Mellon University
Works: 74
Citations: 656
h-index: 15
Badges: Bridge

Publication numbers from OpenAlex, as of 27 October 2026.

Timeline

2014

Internship at Microsoft Research

In 2014, he interns with Microsoft Research’s Software Analysis and Intelligence in Engineering Systems (SAINTES) group in Redmond.
2016

The Apprentice Learner Architecture

In 2016, he coauthors “The Apprentice Learner Architecture: Closing the Loop between Learning Theory and Educational Data.” The architecture models how people acquire cognitive skills from instruction and worked examples.
August 2017

Doctorate in human-computer interaction

In August 2017, he earns his PhD at Carnegie Mellon’s Human-Computer Interaction Institute, advised by Vincent Aleven and Brad Myers. His dissertation, “Projective Replay Analysis,” uses recorded play and an AI player model to evaluate changes to educational games.
By 2018

Systems scientist at CMU

By 2018, he works as a systems scientist at Carnegie Mellon, developing tools that help educational game designers evaluate their designs against their educational goals.
2020

Teaching computers to build tutors

In 2020, he coauthors an AI tutor authoring method with Daniel Weitekamp III and Kenneth Koedinger. A teacher can train the system by demonstrating ways to solve problems and correcting its mistakes; the team tests the method on multicolumn addition.
2024–2025

Dice Adventure for AI teams

In 2024, Dice Adventure serves as the game for a human-AI teaming competition at the IEEE Conference on Games. By 2025, he leads a CMU team developing games for human-AI research; Dice Adventure asks people and AI teammates to coordinate with different information and abilities.
2025

General chair of CHI PLAY

He serves as general chair of ACM CHI PLAY in 2025.
As of October 2026

Associate research professor at CMU

As of October 2026, he is an associate research professor in Carnegie Mellon University’s Human-Computer Interaction Institute. His research areas include games, learning sciences, educational technologies, and human-agent interaction.