Data Engineering for AI Applications

The Data Engineering for AI applications working group has been formed to enable discussion around data engineering practices at AI Institutes. One of the main objectives of this group is to develop a common vocabulary that will increase the findability of the research artifacts coming out of the AI Institutes. The researchers at AI Institutes either actively collect data, use existing publicly available datasets—including synthetically derived datasets—or license the third-party datasets for research and experimentation. In case of data collection or model building activities, thinking early of the interoperability issues in terms of how data that is being collected or model that is being built might complement other existing datasets and models will increase the accessibility, usability, findability, and preservation of the artifacts that come out of the AI Institutes.