Canvas-Integrated Chatbots

Project Summary

The Canvas-Integrated Chatbots project focuses on embedding AI-powered customizeable instructional assistants directly within the learning management system (LMS). By integrating bot-based tools into Canvas, this project enables instructors to provide students with contextualized, course-specific support while maintaining control over content, configuration, and a locally housed database of anonomous transcripts for instructor review and research.

This work addresses a core challenge in scalable education: delivering timely, personalized guidance within the platforms students already use, without adding additional burden to instructors.

Educational Problem Addressed

Students frequently seek clarification on course policies, assignment expectations, and conceptual material outside of class time. In large courses, responding to these questions individually is unsustainable, often leading to delayed feedback, inconsistent responses, or over-reliance on informal peer channels.

At the same time, instructors are rightly cautious about deploying generic AI tools that lack course context or raise concerns around privacy, accuracy, and academic integrity. This project bridges that gap by tightly scoping AI behavior within instructor-defined boundaries and institutional systems.

How AI Is Used

AI chatbots are integrated into Canvas to:

  • Answer student questions using instructor-provided course materials
  • Provide consistent, policy-aligned responses
  • Offer clarifications on assignments, deadlines, and expectations
  • Support conceptual understanding without revealing solutions
  • Reduce repetitive instructor workload related to routine inquiries

Instructor-configured prompts, content sources, and guardrails determine the scope and tone of the chatbot’s responses.

Instructor Experience

Instructors retain full control over:

  • What content the chatbot can access
  • How the chatbot responds to student questions
  • When and where the chatbot is available within Canvas
  • How student interactions are logged and reviewed

The system is designed to function as a teaching assistant that scales instructor presence, rather than an autonomous agent.

Student Experience

Students interact with the chatbot directly within Canvas, receiving:

  • Immediate responses to common questions
  • Clarification grounded in official course materials
  • A consistent source of guidance aligned with instructor intent

This reduces ambiguity and supports self-regulated learning without replacing human support.

Privacy, FERPA, and Institutional Considerations

This project explicitly addresses FERPA and PII concerns by:

  • Limiting data access to course-specific materials
  • Avoiding unnecessary storage of student-identifiable information
  • Providing transparency into data flow and system behavior
  • Supporting locally deployable or institutionally managed configurations

Privacy and compliance considerations are treated as first-class design constraints, not afterthoughts.

Deployment Status

Status: Active development / Classroom use
The system has been deployed in instructional settings and continues to evolve in response to instructor needs, institutional policies, and student usage patterns.

Architecture overview and walkthrough video

Alignment with BOBPE Mission

This project advances the BOBPE mission by enabling scalable, personalized instructional support within authentic educational workflows. By embedding AI directly into the LMS, it amplifies instructor reach while preserving pedagogical intent, institutional constraints, and human-centered teaching.