Introducing WellDB, The World’s First Generative Database

What if your database could generate data in real-time? WellDB does this by using DataFlood models as “tables” that you can “query”. Perform CRUD operations, execute stored procedures, and blend synthetic and real data using the tools you already know such as Compass, Visual Studio, SQL Server Management Studio, and others. You can also pick your flavor of query language. Train DataFlood models with inserts and updates, run queries to generate data, and remove training samples with deletes.

Because WellDB “looks” and “acts” like a database, you can do things like mix production data with DataFlood-generated data, set specific fields to specific values, and perform any other operation you desire. WellDB supports “seed” and “entropy” in queries. By setting a “seed” value you can deterministically generate the same result set for repeatability. With “entropy” you can determine how “random” your results will be. These can be specified in queries and in the underlying DataFlood models.

WellDB supports drag-and-drop, so adding new tables or moving them around in a database schema can be done simply by moving files and folders on disk. This makes WellDB a powerful tool for data modeling. Using The DataFlood Editor you can finely-tune your models and then drag and drop into the folders WellDB is configured to watch and presto! You immediately have queryable “tables”. Imagine being able to design your data model during a (hopefully short) meeting and instantly have synthetic data available for development and testing teams to work with. That is the power of WellDB.

As an introduction to this concept, SmallMinds is pleased to announce the availability of MongTap on Github:

https://github.com/smallmindsco/MongTap/tree/main

MongTap is an MCP server/Claude Desktop extension that can create MongoDB collections backed by DataFlood models. Check out the video below to see it in action!