NUH makes AI prove value before scaling

The hospital screens AI tools for business ownership, process fit, user support and measurable outcomes before moving them beyond pilots.

National University Hospital (NUH) is requiring artificial intelligence projects to demonstrate operational value before they are scaled across the organisation.

Dr Ling Zheng Jye, Chief Medical Informatics Officer and Head of the Kent Ridge Office of Innovation at NUH, said successful pilots can still struggle to scale when organisations focus on technical performance without preparing their operations for adoption.

NUH considers four questions before testing new technology: whether there is a clear business owner, whether it fits existing processes, what support users require, and whether outcomes can be measured.

“The technology is actually the easier part. The change and the workflow — that’s really the harder part,” Ling said.

The hospital has deployed speech-to-text tools for clinical work and large language models to reduce administrative burdens. Scaling them requires training, governance, funding and accountability alongside technical performance.

Staff trust is another requirement because healthcare professionals remain responsible for decisions with direct consequences for patients. NUH has invested in digital literacy and responsible AI practices whilst establishing safeguards and keeping clinicians in control of care decisions.

Ling said AI can summarise information, predict outcomes and make recommendations, but should augment rather than replace healthcare professionals.

NUH is also applying technology to connect different stages of patient care. Instead of waiting for patients to arrive at the hospital, it is seeking to identify who could benefit from earlier intervention and direct them towards appropriate services.

Ling said this approach could apply to other industries with fragmented customer journeys. Businesses need to understand the entire journey and use technology to connect previously disconnected interactions.

For NUH, measurable value comes from creating a smoother end-to-end experience rather than deploying a collection of standalone tools that still leave patients navigating fragmented services.

Ling will speak at the Singapore Business Review Technology Enterprise Summit 2026 on 12 November at the Grand Copthorne Waterfront Hotel in Singapore.