Purdue’s AI Writing Assistant Empowers Students to Strengthen Their Skills
Last updated:
If you’ve watched your child struggle with writing assignments, feeling stuck on how to improve before submitting work for a grade, you’re not alone. Many students miss valuable opportunities to develop their skills because feedback comes too late—after the grade is already assigned. Purdue University has developed an AI-powered tool that changes this equation entirely, giving students a chance to grow their writing abilities before that final submission.
TL;DR
Purdue University developed Charlie, an AI writing assistant embedded in the Circuit peer-review platform, providing rubric-specific feedback before final submission.
Charlie doesn't grade students—instead it identifies areas for improvement, helping students strengthen arguments and address rubric criteria.
The tool emerged from collaboration between instructional technology and faculty, with development beginning in 2019 and continuous refinement with the Purdue Online Writing Lab.
Unlike generic AI tools, Charlie's feedback is specific to each course and instructor's rubric, making guidance immediately applicable.
Interest is growing across campus, with departments including pharmacy and graduate programs exploring adoption.
AI That Guides, Not Grades
Purdue University has launched Charlie, an AI-powered writing assistant embedded within the Circuit peer-review platform. Unlike traditional AI tools that might simply complete assignments for students, Charlie provides formative feedback specifically tailored to an assignment’s rubric—helping students understand where their writing could be stronger before they submit for a grade.
“One of the big concerns about peer review has always been the accuracy of the reviewers,” said Jason Dufair, lead application developer. “So what Circuit brought to the table was a way to calibrate each of the students, who are also the reviewers in later phases of the assignment, against a set of known assignments that the instructor has already graded.”
An important distinction: Charlie doesn’t grade students’ work. Rather, it identifies areas where students can strengthen their arguments and better address rubric criteria before final submission.
Charlie emerged from a collaboration between Purdue’s instructional technology team and faculty members like Lindsay Hamm, assistant teaching professor in sociology, who served as a subject matter expert since development began in 2019. Early iterations could predict scores but offered no actionable guidance. Developers worked with the Purdue Online Writing Lab to train Charlie to provide qualitative, constructive feedback.
“Charlie provides general guidance but will never advise students on specific words to change or actions to take,” explained Casey Wright, lead application developer. “This approach reduces student anxiety and mirrors the kind of feedback they might receive in office hours or in the writing lab.”
Unlike generic AI writing tools, Charlie’s feedback is specific to a particular course or instructor’s rubric—making the guidance relevant and immediately applicable to the assignment at hand.
Author Quote"
Quote: One of the big concerns about peer review has always been the accuracy of the reviewers. So what Circuit brought to the table was a way to calibrate each of the students, who are also the reviewers in later phases of the assignment, against a set of known assignments that the instructor has already graded. Attribution: Jason Dufair, Lead Application Developer, Purdue University
"
Not applicable - no significant bias identified
Developing Skills Before Grades
For parents concerned about their children’s writing development, Charlie represents a shift from评价-based learning to growth-focused feedback. Students receive guidance that helps them identify specific areas for improvement: stronger counterarguments, narrower thesis statements, better evidence integration.
“Now a student can come in and say, ‘Charlie keeps saying my counterargument isn’t very strong’ or ‘Charlie says my thesis statement could be narrower.’ Then I can really home in on what they need to work on for those higher order issues,” Hamm explained.
The quality of feedback depends on the rubric instructors create, and the team has developed built-in templates to make adoption easier. With optional engagement and FERPA-compliant data handling, students and parents can feel confident about privacy while accessing this skill-building support.
Key Takeaways:
1
AI-Powered Writing Feedback: Purdue's Charlie provides rubric-aligned formative feedback to help students strengthen their writing before final submission.
2
Growth Over Grades: Unlike traditional feedback that comes after grading, Charlie helps students develop skills by identifying specific areas for improvement.
3
Expanding Across Campus: Success in sociology courses has sparked interest from pharmacy, graduate programs, and other departments.
A Model for Campus-Wide Writing Support
Charlie has already shown success in sociology courses and is generating interest from other departments including pharmacy and graduate governance courses. The approach exemplifies how AI can be leveraged to support skill development rather than replace human learning.
“My hope for AI—and I know it’s an optimistic hope—is that it can break down more barriers and help people talk to each other more confidently as they have more access to information,” Hamm said.
Instructors interested in using Circuit can log in at peercircuit.org, link courses to Brightspace, and set up assignments with just a few clicks. The tool represents a scalable, supportive approach to writing improvement that could transform how students experience feedback across higher education.
Author Quote"
Quote: Charlie provides general guidance but will never advise students on specific words to change or actions to take. This approach reduces student anxiety and mirrors the kind of feedback they might receive in office hours or in the writing lab. Attribution: Casey Wright, Lead Application Developer, Purdue University
"
Every child deserves opportunities to develop their abilities, and feedback that helps them grow is far more valuable than feedback that merely judges where they are. The system that waits until grades are assigned to provide feedback has it backwards—students need guidance during the learning process, not after it. This is exactly what AI can offer: support that builds skills rather than simply measuring current performance. If you’re looking for ways to help your child develop writing skills through actionable feedback, the Learning Success approach combines the best of AI-assisted guidance with proven strategies for building the foundational skills that make all writing stronger.
Is Your Child Struggling in School?
Get Your FREE Personalized Learning Roadmap
Comprehensive assessment + instant access to research-backed strategies
