Acryl Accelerates "AI Hospital" Push in Line with Government's "AI Basic Medical" Strategy

Image of ACRYL's tech keynote, "ACRYL NOW." Courtesy of ACRYL.

Acryl, a company specializing in artificial intelligence transformation (AX) infrastructure, is accelerating efforts to build “AI hospitals” using its proprietary healthcare artificial intelligence (AI) technology. Its strategy is to expand its healthcare AX business by connecting medical data for AI use while retaining existing hospital information systems, and supporting everything from clinicians’ treatment decisions to hospital operations.

Acryl announced on Oct. 28 that it had unveiled its strategy and related solutions for building AI hospitals by combining a healthcare foundation model, a medical data platform, and AI agent operations technology. The announcement was made through the “Healthcare AX” episode of its tech keynote, “ACRYL NOW.”

The business is being pursued in line with the government’s “AI Basic Medical Strategy.” The government finalized the strategy last August and is expanding AI adoption to reduce gaps in regional, essential, and public healthcare and improve access to medical services. Key initiatives include developing Korean AI for healthcare, building AI-powered intelligent hospital information systems, establishing AI-specialized hospitals in different regions, and expanding public healthcare AI infrastructure.

Acryl presented a healthcare AX solution that integrates medical data, AI models, and operating infrastructure while allowing hospitals to retain their existing systems. Its core technologies are its proprietary healthcare foundation model, “ALLM.H”; medical data platform, “NADIA”; and AI operations platform, “Jonathan.H.”

ALLM.H is a Korean healthcare foundation model trained on domestic medical literature and clinical guidelines, as well as medical laws, pharmaceutical data, and disease data. In Acryl’s own evaluation, it answered all 435 questions on the national medical licensing examination correctly and scored 2,980 correct answers out of 3,009 questions across the national examinations for four professions: doctors, dentists, pharmacists, and nurses. Acryl said the model outperformed the three global AI models used for comparison.

Dr.GAL, a healthcare-specific conversational AI developed based on ALLM.H, provides information on domestic drug approvals and clinical evidence to support healthcare professionals in checking information and making treatment decisions.

The medical data platform NADIA connects electronic medical records (EMRs) and data structures that vary from hospital to hospital, aligning them with medical standards. This enables the creation of an “AI Ready Data” environment that AI can use. Jonathan.H provides integrated management of the entire process for medical AI agents, from development and deployment to monitoring and operations.

Acryl plans to combine these technologies to support AI transformation across entire hospitals, going beyond AI adoption in individual departments. It is also pursuing expansion into physical AI, which goes beyond analyzing medical data and making judgments to carry out tasks in real-world healthcare settings.

The implementation approach is also designed to reduce the burden on hospitals’ existing systems. AI systems can be built within hospitals without replacing existing EMRs, and smaller medical institutions can start with a single server equipped with two GPUs. Another feature is that the systems can be operated on-premises, without sending patient data outside the hospital.

Rather than relying on custom development, Acryl aims to shorten AI implementation time to about four weeks by using a product-based delivery model. The company plans to reduce the adoption burden on regional medical institutions, which tend to have fewer financial resources and specialized IT staff, and expand its supply to regional, essential, and public healthcare.

Acryl has also gained experience applying its solutions in real-world healthcare settings. It currently operates healthcare AI solutions across 17 hospitals nationwide, including 11 tertiary general hospitals, covering more than 10,000 beds. It also holds approvals from the Ministry of Food and Drug Safety for five AI software medical devices.

Acryl is also expanding its business into overseas markets. It has signed a definitive contract to supply a hospital information system to Uzbekistan’s state-owned IT company, UZINFOCOM. As part of the Korean ARPA-H project led by the Ministry of Health and Welfare, Acryl is working on projects related to AI-specialized hospitals and agentic AI for regionally self-sufficient essential healthcare, while also advancing its technology.

Going forward, Acryl plans to expand its hospital AX business, focusing on agentic AI that combines healthcare professionals’ expertise with AI. It is also exploring ways to connect existing AI, which analyzes medical data and supports treatment, with physical AI that can perceive real-world healthcare settings and interact with the physical environment.

“For medical AI to become more widespread, it is important to create a technological environment that allows medical data and expertise to be used securely while retaining existing hospital systems,” said Park Oejin, CEO of Acryl. “Drawing on our proprietary healthcare foundation model, AI infrastructure technology, and experience operating in healthcare settings, we will accelerate the development of AI hospitals and expand our business into regional, essential, and public healthcare in line with the government’s AI Basic Medical Strategy.”

Meanwhile, Acryl uses its tech keynote, “ACRYL NOW,” to showcase its proprietary AI technologies and products through demonstrations. The healthcare episode will be released on Acryl’s official YouTube channel. At its annual conference, “ACRYL NEXT,” to be held on Dec. 9, the company plans to present AI technology achievements in severe diseases, cancer, and rare diseases.

This content was produced with the assistance of AI translation services.