Semantic Strategy
& Architecture
for the AI Era

e-health strategy

e-health strategy

We are in a new era, in which Agentic and LLM AI are rewriting the rules of IT. Modern AI radically improves software economics, while presenting new capabilities of insight generation, partly enabled by the fact that it doesn't care about technical data formats. No domain will feel the revolution more than Health IT (HIT).

Previous HIT strategies that centred on data standards, terminology, and legacy assumptions about how and what solutions should be procured, are now legacy themselves.

Concurrently, investment has poured into AI-based projects in all industries, without understanding its limitations and risks.

A contemporary HIT strategy involves addressing the evolving needs of human healthcare, the needs of medical professionals, and the weaknesses of AI, in order to harness its strengths in a safe way for patients.

openEHR

openEHR Expertise

Thomas Beale is the original architect of the openEHR specifications, including the Reference Model, the Archetype Definition Language (ADL), which is an ISO standard, and an advanced process automation language>. He has presented widely on the innovations of openEHR, and worked with numerous ministries and departments of health, regional healthcare jurisdictions, as well as providers like Intermountain Health (Utah), US Veterans Affairs, and many vendors, and has an unparalleled knowledge of the sector.

How openEHR should be understood today has however significantly changed from when it was created. Ars Semantica brings experience with next generation frameworks and an understanding of LLM AI strengths and weaknesses.

SPLASH platform

SPLASH: a Data Platform for the AI Era

Ars Semantica is developing a next generation open health computing platform — SPLASH — not just for patient data, but for inventory, demographics, encounters, and workflow. The framework integrates ontologies and the domain modelling technology developed in openEHR (archetypes). SPLASH innovations are realisable in today’s openEHR deployments and products.

Standards

Making Standards Cost-effective

openEHR and SPLASH models (including terminology) are upstream models of domain truth. They are used to generate downstream data interoperability standards artifacts, such as HL7 FHIR profiles. These techniques can radically improve the cost of use of such standards.