ML-in-the-database startup MindsDB raises $7.6M seed funding

machine learning

MindsDB, a startup integrating machine learning into existing databases, has announced an investment from Walden Catalyst Ventures, which brings the total it has raised to $7.6 million. The new investor joins YCombinator, OpenOcean (a venture fund launched by the creators of MySQL and MariaDB), SpeedInvest, and the University of California Berkeley SkyDeck fund.

The firm’s platform — also called MindsDB — puts in place a predictive AI layer for existing databases. The aim is to provide anyone with knowledge of standard SQL the ability to apply sophisticated machine learning techniques to the data already stored in the database. This in turn should enable organisations to make data-driven business decisions without needing to become AI developers or experts.

MindsDB works by automating and abstracting machine learning through virtual AI Tables in databases. This is done through SQL, although MindsDB states that MongoDB query syntax is also supported.

Once linked to a database, users can enable MindsDB to learn from historical data automatically by training a predictor using a single SQL statement. Predictions can then be made by querying MindsDB virtual AI Tables, which behave the same way as normal database tables. Results can then be fed through to forecasts in BI dashboards, all through standard SQL, according to MindsDB.

Databases supported by MindsDB include MongoDB, PostgreSQL, MariaDB, MySQL and Microsoft SQL Server.

The firm also confirmed partnerships with cloud database developers including Snowflake, SingleStore and DataStax, to connect MindsDB’s machine learning layer to their platforms. Users of those products will therefore be able to take advantage of machine learning capabilities provided by MindsDB from within those platforms, turning their databases into predictive engines, the firm said.