Synapxe

Synapxe's role is highly unusual: it is not a hospital, not a research institute, but the "shared IT department" for all public healthcare institutions. This means:

• Unified data platform: every public hospital's electronic health records flow into one national system (NEHR), so AI models can be trained on nationwide data
• In-house AI products: Synapxe doesn't only buy commercial AI — it builds its own AI tools spanning screening, imaging, and administrative workflows
• Unified deployment: once an AI model is validated, it can be rolled out to every public hospital simultaneously

Representative AI products:

• ACE-AI: an AI health-screening tool predicting diabetes and hyperlipidaemia risk, scheduled to roll out to all Healthier SG clinics from 2027
• Clinical Note Summarizer: an LLM-based EMR summary tool
• Imaging AI: AI-assisted radiology diagnosis (jointly developed with NUH and SGH)
• Administrative automation: AI automation for prescription handling and insurance claims

Synapxe collaborates extensively with AI Singapore, NUS Medicine, and the public hospitals (NUH, SGH, TTSH, and others), making it the central node of national medical AI.

Synapxe's core innovation in medical AI is "national-grade data + national-grade deployment".

The two biggest pain points facing most medical AI startups globally are:

• Data: medical data is scattered across different hospitals and EMR systems, making large-scale training hard
• Deployment: every hospital's IT systems and compliance processes are different, so a single product takes months to land at one site

Synapxe's institutional setup dissolves both problems: it directly owns nationally unified medical data (NEHR), and once an AI tool is built it can deploy simultaneously to 46 hospitals and 1,400+ clinics. This "national-grade data + national-grade deployment" advantage is rare globally — only the UK's NHS and Denmark's healthcare system come close.

Technically, Synapxe's AI line is pragmatic:

• Doesn't chase frontier model architectures
• Prioritises deployment reliability, compliance, and explainability
• Uses a lot of "AI + human-in-the-loop" hybrid workflows
• Cautious on LLM applications (currently mostly in low-risk scenarios like clinical-note summarisation and form handling)

In Singapore's AI strategy, Synapxe is "the model for sectoral AI deployment" — healthcare is the only sector where "national-grade AI infrastructure + national-grade data + national-grade deployment" are fully connected.

In the "seven transmission levers" framework:

• Lever 3 (industry adoption): the execution body for medical AI deployment
• Lever 5 (government adoption): AI-enabling the public healthcare system is the largest single government-adoption scenario

A take: Synapxe's institutional setup is Singapore's natural "small country" advantage cashed in — a population of just 5.8 million, high concentration of the national healthcare system, and a unified IT architecture. This lets it do what large countries like the US and Japan cannot: train national-grade models on national-grade data and deploy them to serve the entire population.

ACE-AI is the most representative case: train a model on nationwide diabetes / cholesterol screening data and deploy it to all Healthier SG clinics by 2027 — in theory benefiting all 5.8 million residents. Medical AI deployment at this scale only works under "small country + centralised system".

But Synapxe also faces challenges: medical AI's safety-cautious requirements make it iterate slower than commercial AI, its boundary with commercial AI companies needs clarification (in-house vs procured), and its international data-governance cooperation is constrained by PDPA.