Optimize session control and generation quality by attaching or detaching execution context for your AI chains with fast and reliable CDB service.
CDB, or Context Storage, stands as a highly available and exceptionally fast-to-read storage service tailored for context retention across multiple executions within your application. Whether it's preserving session context in a chatbot or tracking historical runs within a single chain, CDB offers versatile options for context management.
With CDB, you can effortlessly attach context to any execution as needed. You retain control for manual search across contexts, or opt for the convenience of C4AI's automated context handling.
During the creation process, CDB presents a range of configurable parameters. These include storage size, strategy for receiving embeddings, scalability options, and retention policies, among other valuable settings.
CDB seamlessly integrates with AIL, AAC, and even your own custom models. This integration not only streamlines context management but also saves development time by eliminating the need for additional storage solutions with specific retention policies.
Empower your application with CDB's efficient context storage capabilities, designed to enhance performance, scalability, and overall ease of management.
The service assumes that you are using it for short-term operations and therefore high data availability can be ensured up to 24 hours from the moment of the first recording.
However, in the settings you can select additional storage options that will allow you to keep the context indefinitely and use high availability if necessary.
For high-quality content, this is up to 1GB of RAM. However, you can use long-term storage to significantly increase this volume. You can also choose your own strategy for updating highly available data.
The storage itself is intended for text data or vectors. However, you can store any other files linked to their location.