Before AESOP became a standalone company, its journey began over a decade ago at Taipei Medical University as a long-term research project led by Prof. Yu-Chuan Li, whose goal was to make medication use safer. Together, Li and his team laid the foundation for what would later become AESOP. Over time, they expanded the project’s data resources and analytical capabilities through years of research and collaboration.

​

In those early days, computing power was a far cry from what it is today. When Li and his team first tried to understand the relationships between diseases and medications, they compiled 12 years of health insurance data covering 7,000 diseases and 1,400 medications. They also used computational resources from the National Center for High-Performance Computing. However, they were told that analyzing such a massive dataset would take more than 100 years.

​

To address this challenge, they developed a more practical approach. Rather than analyze the entire dataset at once, they first examined relationships within each data dimension, beginning with relationships among diseases and then applying the same method to relationships among medications. They then combined the results and extended the analysis to examine relationships between diseases and medications.

​

As the research moved toward clinical application, it revealed a new set of challenges.

​

The research team soon realized that even groundbreaking work would have limited impact unless it could be developed into viable, scalable products. Only then could they translate their research into clinical practice and help physicians address real clinical problems. They also faced a greater challenge: the complexity of electronic health record systems, which far exceeded what they had anticipated.

​

To bridge the gap between research and real-world implementation, Li formed an entrepreneurial team and invited Dr. Jim Long and Dr. Jeremiah Scholl to help turn the research into practical solutions that could improve healthcare.

​

Since spinning off as an independent company in 2019, AESOP has rapidly expanded its medical records database, integrating Taiwan’s National Health Insurance data with large datasets from U.S. Medicare, Mayo Clinic, and DARTNet. Its database soon grew to include more than 3.2 billion medical records, providing the foundation for further development.

​

Advances in computing have enabled AESOP’s AI to analyze these records and extract more than 5 million clinical decision pathways. The result is Clinical Decision Intelligence, a new class of medical AI designed to reconstruct, model, and analyze clinical decision-making across diagnosis, treatment, medication management, and medical coding. By examining how and why decisions are made rather than predicting outcomes alone, this approach provides more explainable, context-aware, and traceable decision support, helping users understand and evaluate the evidence behind recommendations and increasing trust in the decision support they receive.

​

As AI capabilities continue to advance, AESOP remains focused on using technology to improve patient safety and quality. Through flexible solutions that support the meaningful use of medical AI in clinical decision-making, AESOP is working toward a healthcare system in which data flows more effectively and physicians can make better-informed decisions.