AI Technical Team
The AI Technical Team is one of the main development units responsible for engineering applied AI solutions across campus. Our primary purpose is to transform teaching, learning and operational efficiency by building custom applications, intelligent agents and conversational assistants. We partner directly with faculty, students and staff to leverage AI for specific, high-impact use cases that enhance the academic experience and streamline campus operations.
Enhancing Teaching and Learning
Our goal is to integrate AI tools that directly support faculty instruction and student success.
We have developed NotebookLM Teaching Assistants that use course materials to help students study and review, and a Grading Assistant that uses transcripts and rubrics to make skill assessment grading more efficient and consistent for faculty. We are building a Scheduling Assistant to help students plan study blocks based on their assignments and calendar, and we are developing practical NotebookLM examples and documentation to inspire broader academic adoption of AI tools.
Streamlining Campus Operations
We have implemented several AI-driven solutions to improve administrative processes and enhance support across campus, including:
- Built an IRB application assistant that provides structured feedback on submitted research proposals.
- Deployed the AI Website Assistant and initiated development of a General Support Chatbot powered by the Appalachian Technology Knowledge Base to answer common IT questions.
- Applied AI/ML in Campus Dining to support labor optimization and improve operational efficiency.
Planned projects include an Admissions Assistant to help prospective students navigate the application process, and several HR use cases leveraging Copilot Studio agents to assist recruiting teams.
Foundational Services and Secure Development
We provide the secure infrastructure and tools necessary for the entire campus to experiment with and deploy AI safely.
We are building out secure, flexible platforms, including custom Gemini examples and documentation for the community and a Confluence-Copilot Connector to leverage existing knowledge sources. We are also building out an on-premise model hosting solution.
Live Projects
AI & Support Website Assistants that provide real-time answers for the official campus AI and ITS Support websites. Both knowledge bases are restricted to public university documentation.
The IRB AI Proofreading Tool uses Azure OpenAI to refine your Cayuse IRB applications before submission. Focuses on formatting; does not replace the review of the IRB committee.
Custom NotebookLM environments tailored for specific academic and administrative roles like analyzing complex materials, building interactive notebooks and automating course scheduling.
Software Catalog AI Analyzer designed to map active features and cut costs on redundant tools. Provides centralized data for faster collaboration while leaving final procurement decisions to ITS leadership.
NotebookLM Demo
Learn how to use NotebookLM
to create a study schedule that can be imported into your Google Calendar.
For detailed instructions and resources, follow the knowledge base article NotebookLM: Scheduling Assistant.
Projects in Development
AI Phone Agent for Canvas Support: An Azure-based voice agent for after-hours Canvas LMS help.
- Provides 24/7 tech support using official documentation
- Reduces routine help desk tickets through automated self-service.
Programmed to escalate complex issues to human staff.
Enrollment & Admissions Website Assistant: A unified chatbot for Admissions, Financial Aid and Graduate Programs
- Consolidates institutional data for all student types, bridging the gap between automated help and personal outreach.
- Generates leads and tracks engagement for Slate integration, alerting staff when to contact prospective students.
AI for initial engagement only; a human recruiter initiates the final outreach.
VetTech Grading Assistant
Uses AI to analyze video transcripts against rubrics for practical skill assessments.
- Dramatically reduces faculty time spent grading hundreds of hours of video assignments.
- Automatically flags rubric milestones and identifies content gaps.
- Currently undergoing pilot testing to ensure it meets specific pedagogical needs.
AI flags rubric milestones in transcripts; faculty maintain sole authority over final skill validation.
Campus Dining Labor Optimization
Employs machine learning to optimize staffing levels by analyzing transactional and labor patterns.
- Aligns staff schedules with predicted peak hours to optimize labor.
- Reduces wait times for students and staff during high-traffic intervals.
- Moving from data-cleaning into fall testing for predictive model deployment.
The model provides recommendations based on historical data; human managers retain control over schedules.
Gemini Gems Library
A suite of specialized AI Gems tailored for faculty, staff, and students. All Gems operate within a "Human-in-the-Loop" framework; users are prompted to verify AI outputs.
Career Architect
A specialized coach that prepares students for the workforce by:
- Matching resume gaps and student skills to job market demands while keeping the student actively involved.
- and conducting mobile-optimized, voice-driven mock interviews.
Critical Thinking Partner
A neutral Socratic coach that pushes student thinking further without providing final answers by:
- Maps arguments to proven learning frameworks to quickly spot student bias and logical fallacies.
- Restricts the AI from generating final assignment text
Canvas Quiz Assistant
Automatically generates high-quality quiz questions formatted specifically for Canvas import.
- Breaks down source text to cover every section and exports ready-to-use Canvas XML without errors.
- Formats complex math and chemistry formulas accurately using subject-aware LaTeX rendering
Active Learning Architect
Transforms traditional lectures into engaging, student-centered learning experiences.
- Structures sessions around the "10-2 Rule" to prevent overload with backup plans for low engagement.
- Spots tough "stick points" throughout the lesson to catch student confusion early.
Training Video Assistant
Converts screen recordings and visual demonstrations into structured, written training guides.
- "Watches" user-uploaded videos or YouTube links to accurately document each step.
- Outputs final documentation in multiple popular, easy-to-use formats.
Feedback Parser
Transforms raw survey data and comments into understandable and actionable key insights.
- Scores feedback impact on a 1–10 scale and pinpoints extreme reviews for valuable lessons.
- Identifies recurring keywords to map thematic skills, strengths, and hidden growth opportunities.