The US healthcare system generates millions of medical records, healthcare payments, and related administrative data points every minute. Significant efforts to standardize formats, create clearinghouses, and apply interoperability tools have all failed to connect these datasets in a meaningful way. There is tremendous value in connecting these datasets and discovering the insights across them.
Given the variation and scale of data, this continued challenge can only be met by a solution that can learn and evolve on its own. The recent breakthroughs in machine learning and big data have created the opportunity to develop a practical, long-term solution. Standardizing and transforming different data sources through automated ML solutions that are architected to be source-agnostic, scale to billions of records, and to natively check quality and completeness will finally make interoperability a reality.