DataStax Inc. today introduced new versions of Langflow and RAGStack, two software tools designed to ease the development of artificial intelligence applications.

Santa Clara, California-based DataStax is backed by more than $300 million in funding. Its flagship product is a cloud-based NoSQL database called Astra DB that is available in two editions. There’s a standard edition and an AI-optimized version capable of storing vectors, the mathematical structures in which neural networks keep the information they ingest.

“The Generative AI stack is a big and complex ball of technology that many are working to get their arms around,” said DataStax Chief Product Officer Ed Anuff. “We’re focused on helping developers stay true to their roots so they can do what they do best: build and develop, rather than worrying about application infrastructure.”

Langflow, the first offering that DataStax has updated, is an open-source tool for developing AI applications. It enables software teams to perform many of the tasks involved in building an AI service through a drag and drop interface. The tool thereby reduces the need for custom code, which speeds up development. 

The new Langflow 1.0 release that DataStax debuted today adds a set of AI memory management features. According to the company, the enhancements will make it easier for applications such as chatbots to process text prompts entered by users. Another new capability included in the update will enable developers to build multimodal AI applications, or programs that can process not only text but also images and other types of input.

A second set of feature additions is designed to boost software teams’ productivity. According to DataStax, Langflow 1.0 includes new project examples for first-time users and a preview feature that makes it easier to evaluate the accuracy of an AI application’s output. In conjunction, the company is launching a free cloud version of Langflow that will allow developers to test the tool without downloading it on their machines.

The other product that DataStax updated today is RAGStack. It’s a collection of software building blocks for building AI applications with RAG, or retrieval-augmented generation, features. It includes Langflow, open-source machine AI models and several other components. 

RAGStack 1.0, the new version of the offering that DataStax debuted today, includes a new feature dubbed Knowledge Graph RAG. It provides the ability to store the information used by an AI model as a graph, a data structure that highlights when two pieces of information are connected to one another in some way. That makes it easier for neural networks to spot useful patterns in the data they’re analyzing.

RAGStack 1.0 also includes two specialized AI models called ColBERT and Text2SQL. The former model is primarily used for search tasks: It can run queries on large text datasets in tens of milliseconds. Text2SQL, in turn, allows users to query SQL databases with natural language prompts.

Image: DataStax