It is a constant frustration for those who wish to become FAIR that the FAIR Principles provide no guidance on how they should be achieved. A wide range of “FAIR enabling technologies” are available, and there has been some nicely grounded effort to create subsets of technologies that work well together within specific communities (e.g. the FAIR Implementation Profiles work from GO-FAIR). However, the problem still remains: How do I know if I am FAIR? Attempts to measure FAIRness have now become a cottage industry, with ~30 independently-authored FAIR assessment tools being publicly available. Unfortunately, these tools frequently provide dramatically different assessment outcomes, leaving one to wonder which (if any) is correct. I will discuss the problems associated with FAIR Assessment, and explain the path that is being taken by the OSTrails EOSC Infrastructure project to attempt to harmonize FAIR assessment for all communities. I will then discuss an example of a research infrastructure in the domain of biodiversity conservation where the infrastructure has been designed to be FAIR at-birth. Environments where every resource is reliably and deeply FAIR results in the ability to automatically generate unusually powerful discovery and analytics environments, where much of the complexity of data discovery, integration, and even selection of analytical tools, can be hidden by the rich metadata annotations of a FAIR resource.