Introduction
Miami University is introducing the LLM Pilot Platform, a secure, university-managed gateway to generative AI. Through a single, familiar chat interface, students, faculty, and staff can work with a range of leading AI models — the same class of models behind today's popular assistants — without needing their own paid subscriptions or personal accounts. The service runs entirely within the university's own secure AWS environment, so conversations and institutional data stay under our control and are not used to train outside vendors' models. Because the gateway connects to multiple model providers at once, we can offer choice, add new models as they emerge, and avoid being locked into any single vendor, all behind one consistent, university-supported front door.
How to sign on
- Visit the web front end at https://openwebui.itapps.miamioh.edu/
- Click the "Continue with Microsoft" icon
- Login with your MiamiOH credentials
The value it brings
The benefit looks a little different for each member of our community. Students gain equitable, university-provided access to capable AI tools for learning, studying, and building skills they will carry into the workforce, regardless of whether they could afford a commercial subscription on their own. Faculty gain an institutionally approved tool they can bring into the classroom and their research with confidence, including the ability to request access for a specific course and have usage scoped appropriately to it. Staff can use it to speed up every day administrative, analytical, and communications work. And for the broader university, the gateway provides the guardrails that make broad access responsible: Per-group budgets, usage visibility, control over which models are available, and strong data protection, so we can widen access with confidence while managing cost and risk centrally.
Eligibility
Access to the pilot platform is by invitation only.
Constraints and Boundaries
Each person participating in the pilot will be part of a token allocation / budget. Consumption and budget remaining are displayed as you consume tokens.
Help and Support
Part of the purpose of this pilot is to learn what support is required and how we can efficiently connect people with the resources they need.
Use this support request form to initiate the request. Example use cases we envision:
- Enabling a specific model
- Adjusting budgets
- Using API keys