Developing a National AI Virtual Research Environment (VRE) Infrastructure for Healthcare Analytics

The Challenge

Healthcare research in Australia increasingly depends on advanced analytics and artificial intelligence (AI). However, researchers face significant infrastructure barriers.

Many healthcare researchers are not software engineers. Setting up AI environments, installing complex software stacks, or accessing GPUs can be time-consuming and technically difficult. As a result:

• research teams duplicate infrastructure efforts across institutions
• sensitive data cannot easily be analysed in secure environments
• access to high-performance compute is limited
• AI experimentation is slowed by fragmented systems.

Commercial cloud platforms are costly and raise data sovereignty concerns. There is currently no nationally coordinated, secure and scalable AI-enabled virtual research environment (VRE) tailored to healthcare research needs. Without such infrastructure, the pace of AI-enabled health research is limited.

The Response

This project is developing a national AI virtual research environment (VRE) providing researchers with secure, scalable access to advanced analytics and AI tools.

Built on the ARDC Nectar Research Cloud, the VRE will provide a portable and reproducible research platform that removes the need for individual teams to build and manage complex infrastructure.

Phase 1 will deliver a minimum viable product (MVP) that establishes:

• VRE environments for working with public datasets and training for early-career researchers
• secure VRE environments aligned with ISO 27001 principles for handling moderately sensitive healthcare data
• a customised microscopy VRE environment designed for large imaging datasets and domain-specific workflows.

The platform will be designed for expansion. It will establish the foundation for future trusted research environments (TREs), federated learning capabilities and broader national deployment.

Who Will Benefit?

The Partners

This project is delivered through a co-investment partnership between:

The University of Queensland

The University of Queensland (UQ) is the lead contracting organisation. UQ is responsible for overall project leadership, architecture design and core development of theVRE.

Intersect Australia

Intersect Australia is a subcontractor and deployment partner. Intersect contributes expertise in data science, AI integration and scalable infrastructure deployment.

ARDC

ARDC is a co-investor in the project and provides national infrastructure coordination, program oversight and alignment with the People Research Data Commons. The project also works with the Microscopy Australia community through a dedicated work package to develop a specialised VRE tailored to microscopy research needs.

Target Outcomes

Phase 1 will deliver:

• a minimum viable product (MVP) that is a national AI VRE deployed on at least 2 nodes of the ARDC Nectar Research Cloud (Nectar)
• security controls designed to align with ISO 27001 principles, supporting moderately sensitive healthcare data and establishing the basis for future certification pathways
• integrated GPU-enabled computing environments for AI and advanced analytics workloads
• a user-friendly virtual desktop environment to simplify access to tools and workflows
• foundational metadata and resource discovery capabilities designed to support future persistent identifier (PID) integration
• a specialised microscopy VRE, initially deployed on selected Nectar nodes as part of the MVP
• a sector engagement and partnership roadmap to inform future Phase 2 development.