GenAI for harmonizing common data elements (CDEs).

Common data elements (CDEs) and data dictionaries are the core of data harmonization. We’ve been working hard to expand the capabilities of our DIVER tool for data discovery and automated data harmonization.

CDE generation and harmonization across data silos can be painful and really increase the activation energy needed for analysis and innovation.

We leveraged generative AI to rapidly create a harmonized list of over 14K CDEs across brain and age focused studies as sparse and heterogenous datasets. This has saved months of analyst time! We hope to expand this to other research domains in the very near future. Shout out to Alan Long from our team for curating this effort and leading the development.

Please see our presentation on the DIVER platform and our development of GenerativeCDEs in this presentation we did at NIH yesterday. Slides linked here.

“One data dictionary to rule them all!”