AI Assistants

Problem

Generic AI chatbots cannot reliably answer questions using University‑specific information and pose risks around inappropriate use or exposure of institutional data.

Delivery

The Innovation team proactively developed a platform for creating AI assistants using retrieval‑augmented generation (RAG), enabling chatbots to provide responses grounded in approved University content, and trialled the approach across use cases including AI teaching assistants and a Careers Network chatbot.

Outcome

The work demonstrated that secure, data‑grounded AI assistants could be delivered safely and at scale, and was fundamental in prompting the University to progress its use of AI chatbots, directly leading to the adoption of NebulaONE as its strategic platform for AI assistants.