Knowledge Sources

Knowledge Sources enhance the AI Chat Assistant's capabilities by connecting it to a variety of external data sources and uploaded files. These sources, whether structured (such as databases, CRMs, or APIs) or unstructured (like documents, PDFs, and websites), provide the context necessary for the AI to deliver highly accurate and relevant responses.

When a user interacts with the AI, it pulls data from the connected Knowledge Sources in real-time, ensuring that responses are informed by your business-critical data, documentation, and other relevant resources. This integration empowers the AI to provide context-aware answers, backed by up-to-date information.

Available Integrations

File Upload

Directly upload documents for agents to reference

Box

Access enterprise content managed in Box

Dropbox

Access enterprise content managed in Dropbox

Google Drive

Access files and documents in GDrive

Notion

Access documents and data in Notion

OneDrive

Access content stored in the OneDrive cloud

Web Scraper

Pull data from any internal or external website

Confluence

Access Confluence documentation shared by your teams

Configuring Knowledge Sources

To configure Knowledge Sources, navigate to the Appsmith AI Datasource section within the application. From this interface, you can import your desired data sources or upload files for integration.

You can further customize the interaction between the AI and your data by adjusting the following configuration options:

• Chunk Size: Define the size of the data chunks the AI processes. This setting allows for more efficient data management and enhances performance when handling large datasets.

• Chunk Overlap: Adjust the amount of overlap between data chunks to ensure continuity and context when the AI processes extensive documents or datasets.

These configuration options allow you to optimize data parsing, ensuring the AI delivers more accurate and contextually relevant responses.