How to Use Knowledge Bases

What is a Knowledge Base?

A knowledge base is a centralized repository of information, documents, or data that can be searched and referenced to answer questions, solve problems, or provide context.

StackAI enables users to leverage a powerful and flexible RAG (Retrieval-Augmented Generation) system through a simple drag-and-drop interface. By connecting directly to their knowledge base, users can effortlessly incorporate contextual search capabilities.

To use a Knowledge Base in your project, drag and drop a Knowledge Base Node or select the Knowledge Base tool in your LLM. Knowledge Bases can be created and managed in the Knowledge Base Dashboard, or you can create one on the fly in your Knowledge Base Node.

The Knowledge Base Node

The Knowledge Base node acts like a search engine over files, allowing LLMs to retrieve the precise context needed to perform any given task effectively. Indexing and storage in vector databases happen automatically, without any action from the user. Syncing also occurs automatically, if enabled, so that new files or changes are added to the knowledge base.

Unlike other platforms, StackAI enables users to leverage a powerful and flexible RAG (Retrieval-Augmented Generation) system through a simple drag-and-drop interface. By connecting directly to their knowledge base, users can effortlessly incorporate contextual search capabilities. The Knowledge Base node acts like a search engine over files, allowing LLMs to retrieve the precise context needed to perform any given task effectively. Indexing and storage in vector databases happen without any action from the user. Syncing occurs automatically if enabled, so that new files or changes are added to the knowledge base.

Import Knowledge From Any Source

StackAI allows you to create a knowledge base from uploaded documents, tables, Dropbox, Google Drive, Sharepoint, and many more! File metadata from platforms like SharePoint is also imported to enhance information retrieval, resulting in a 27% increase in accuracy for financial applications.

Advanced search controls within the knowledge base, designed with the right level of abstraction to ensure ease of use for end users.

Citations are clearly displayed in the user interface to enable response auditing. Users can view the original source file and the exact information chunks utilized by the LLM. Superscript references within the response allow users to easily trace and verify the underlying data.