Natural language processing is the subfield of AI concerned with training models against human language to generate insights, sometimes in the form of conversational AI. Examples of NLP might be extracting recommendations from text documents, or training context specific chat bots for delivering customer support or follow up.
Data used in NLP is often tabular, with mappings of "intents" to phrases, and other corpuses of text data. This kind of data fits naturally into a relational model, and we have seen strong interest from NLP practitioners in leveraging Dolt's combination of SQL and version control in their infrastructure.
Kalido, a company building AI tools based on customer natural language data, uses Dolt in their model delivery pipeline. You can read more about how Kalido uses Dolt here.