Healthcare data platform company Lulumedic said on May 18 it has built its AI and cloud-based next-generation medical mydata service on Amazon Web Services (AWS).

The company said demand has recently been rising in South Korea for “medical mydata” that allows patients to take the lead in consolidating and managing scattered personal medical records.

But security concerns and stringent compliance requirements have acted as obstacles to adopting AI when linking sensitive personal data such as disease history, genetic information and mental health records with external AI models.

Lulumedic said it set up a closed network environment isolated from outside using Amazon Virtual Private Cloud (Amazon VPC), which provides an independent virtual network space within the AWS Seoul Region, so that all workloads and patient data remain separated from external access. It said this blocks external intrusion during data transmission and processing and ensures a data sovereignty framework in which patient data is processed in South Korea.

Lulumedic also uses AWS machine learning infrastructure by training and fine-tuning a healthcare-focused large language model (LLM) called TrustLLM, jointly developed with AWS partner Upstage, on Amazon SageMaker AI, and running inference on Amazon EC2.

It said it built a “Dedicated Zone” on Amazon VPC that operates independently without external data leakage and implemented a “stateless architecture” in which patient data is not stored or reused for AI training during inference. It stressed that this technically meets the de-identification and non-retention principles for personal data required by South Korea’s Medical Service Act and supports medical staff in using generative AI safely in real time.

Lulumedic CEO Young-woong Kim (김영웅) said the success of medical AI depends not only on performance but also on how perfectly patient information can be protected. He said Lulumedic is proving the effectiveness of data use through its medical mydata service, Distat, and will combine AWS security infrastructure with Upstage’s AI capabilities to build the safest and most innovative Korean-style medical data ecosystem that overcomes regulatory barriers.

Jeong-won Yoon (윤정원), head of the public sector at AWS Korea, said the launch of the Lulumedic platform will be a starting point for South Korea’s medical industry to shift its data use paradigm from “institution-centered” to “patient-centered,” going beyond the achievement of a single company. He said it is also meaningful in that it presented a best-practice case of thoroughly complying with South Korea’s data sovereignty and regulatory requirements by safely storing and processing patient data domestically based on the AWS Seoul Region.