DataStax’s Astra DB database has received an extension for the Google Vertex AI platform. This allows users to load database data directly into the AI platform.
With the newly released Astra DB Vertex Extension for Google Cloud’s Vertex AI Platform, it is possible to directly load data from the Astra DB cloud-based database into Vertex AI. This, according to the provider, should ensure that the development of LLM models and generative AI solutions with the Google platform can be faster, smarter and more relevant.
Vector-based data
The underlying Astra DB cloud-based database is built on the open-source Apache Cassandra database. Through the extension the company has shown, Astra DB feeds the Google Vertex AI environment with so-called vector-based data.
Using the extension and the Astra DB database makes it easier for companies to supplement the standard data on which Vertex AI models and generative AI solutions are trained with their own data. Gaps in the existing data are thus more easily filled that can help at the exact moment when a search query retrieves the enhanced information.
In addition, this can reduce AI hallucinations, DataStax indicates. The database extension is now available immediately.
Tip: MongoDB Queryable Encryption protects sensitive data, even when it’s in use