About the Project
An academic AI initiative — owned by SCSU, built for academia, free for everyone.
HuskyBot is the St. Cloud State University Library's open-source AI assistant. We're combining a library chatbot with a research-writing assistant into a single, custom-trained tool — deployable both on the web and as a downloadable model.
Our Mission
Build a purpose-trained academic AI that helps SCSU students do better research, write stronger papers, and cite more confidently — while modernizing how the library serves its community.
Our Vision
A future where every academic library can run its own trustworthy AI — owned, controlled, and accountable to the community it serves. HuskyBot is our blueprint for that future.
Three Goals, One Tool
What HuskyBot is being built to do.
A Library Chatbot
Answer questions about SCSU library resources — hours, services, databases, research guides, and more — 24/7, with the precision of a reference librarian.
A Research Writing Assistant
Help students draft and refine academic papers with citation generation, literature review support, and academic-tone checking.
A Custom Academic LLM
A fine-tuned open-weight model trained on academic writing styles and scholarly literature — not a generic, off-the-shelf chatbot.
Dual Deployment Model
Use it on campus. Or take it home.
HuskyBot is built to be available everywhere SCSU community members work — online and offline.
Web-Accessible
Embedded directly into the SCSU library website and available as a standalone web app. Open it in any browser — no install required, no account hassle.
- Always up-to-date with library resources
- Works on any device
- Single sign-on with SCSU credentials (planned)
Downloadable
A packaged GGUF version runs on your own laptop or workstation — entirely offline. Compatible with Ollama, llama.cpp, and standard local LLM tooling.
- No internet required after install
- Your data never leaves your device
- Works on consumer-grade hardware
What We Stand For
The principles behind HuskyBot.
Open Source
Code, weights, and documentation are public. Other universities can fork, adapt, and improve.
Privacy-First
Runs on SCSU infrastructure. Conversations stay on campus. No third-party data harvesting.
Built by Students, for Students
Led by SCSU faculty and students — the people who actually use libraries every day.
Reproducible
Documented end-to-end so other institutions can stand up their own academic AI.
Ready to dig deeper?
Explore the technical approach, the project timeline, and how we're building HuskyBot lean.