SuperDuperDB, an Intel Ignite portfolio company, has released version 0.1 of its open-source framework designed to streamline the development and deployment of AI applications. This Python package enables the integration of various AI capabilities, including machine learning models and AI application programming interfaces, with existing databases, allowing users to build AI applications directly on these databases. The framework supports a wide range of popular AI models and databases and has garnered $1.75 million in early funding from investors like Hetz Ventures, Session.vc, and MongoDB's venture capital arm.
SuperDuperDB addresses the complexities faced in deploying machine learning models and proprietary data in business applications. Traditionally, this process involves intricate and time-consuming data extraction and transfer between main and specialized vector databases. SuperDuperDB's solution involves embedding AI models, including streaming inference and scalable model training, directly within the enterprise's existing database. This approach simplifies the process, allowing developers to deploy a scalable environment for AI models and APIs that communicate directly with the database. This environment can be set up in various modes, including experimental, single client, or scalable compute in cloud or on-premise settings. SuperDuperDB's offering not only simplifies the use of standard machine learning models but also facilitates the integration of advanced generative AI models and custom models for specific use cases. The product, though relatively new, has already seen significant interest from major players in the industry, supporting a broad range of databases and AI models, and is in talks for further expansion of its ecosystem.