The AI Services Framework provides a faster and more secure way to bring AI solutions from idea to reality.
AI activity is accelerating across New South Wales. From chatbots for students to analytical tools and multi-model applications, teams across the university are building AI solutions that connect with UNSW’s data and applications.
What is an AI solution?
AI solutions include not only the general or personal use of AI, but also building, integrating or deploying AI capabilities within UNSW’s systems or processes.
Possible AI solutions include:
- Chatbots or agents connected to UNSW data or systems
- AI tools embedded in business processes or platforms
- Models that make or support decisions
- AI services built for staff, students, and researchers.
AI solutions are not:
- Personal or experimental use of approved AI tools
- Using ChatGPT or Copilot for individual tasks such as drafting text or analyzing documents.
Until recently, UNSW teams building AI solutions have had to research which type of AI system best suits their needs. Universities have design and governance processes, but they were not built with AI in mind. Without AI-specific patterns and guidance, projects were slow to align AI solutions with existing architectures and review requirements, often duplicating efforts along the way.
framework
of AI Services Framework (AISF) We’re changing that. Joint development with AccentureAISF provides teams with a structured, reusable way to design and review AI solutions with safety, privacy, and governance guidance built in from the beginning. This framework integrates three core components:
- Design patterns that show how to build different types of AI solutions
- Reference architecture mapping standard technical building blocks
- A functional model that defines the common language needed to design, deliver, and manage AI solutions.
Three design patterns are available covering single-agent systems, multi-agent orchestration, and third-party tool integration, with more development and work underway to keep it up to date.
Rather than starting with a blank page, teams choose patterns that meet their needs and build from there. For low-risk applications such as AI recommendation systems or personalized content generation, these design patterns can reduce build times from 10-12 weeks to 4-6 weeks.
actual proof
project scouta 24/7 AI assistant designed to help students navigate registration, timetable, and administrative processes, was the first effort to use AISF end-to-end. Rather than building everything from scratch, the Scout team started with a framework template and tailored it to their needs.
The design and documentation effort was reduced by about half, and a significant portion of the solution documentation was drawn directly from pre-built templates. This project completed the enterprise architecture review without requiring any rework. This process can add several weeks to your timeline. Scout then received architectural review committee approval and proceeded with delivery.
living framework
AISF is currently public and available, but it is designed to evolve. Lessons from early adopters like Scout are already feeding back into the next iteration of patterns and guidance. The resources will continue to improve as more projects put the framework into practice. Feedback from our community will shape what comes next, and we welcome contributions from across the university.
Whether you’re proposing an AI solution or are simply curious about how UNSW approaches delivering AI, you can explore the AI Hub framework. AI service framework page.
