Overview

Create and manage data stores—a collection of various data sources like PDFs, webpage links, or knowledge corpora—through the organization console.

Data stores created at the organization level have a global scope, accessible by all projects and bots, while project-scoped data stores are restricted to specific projects or bots within that project. Reusable data stores help you avoid duplication tasks of creating data sources, save costs, and effortlessly integrate across multiple AI bots or other AI-powered applications. One setup, multiple uses!

• To access and manage data store details, follow these steps and refer below image:

• Select a Data Store or Click on any data store to open its details.

• Navigate to Overview (1): In the overview section, you will find detailed information (2), including the data store’s name, description, and the AI bots utilizing it.

• Access RAG Settings (3): At the bottom of the overview section, you will find the RAG (Retrieval-Augmented Generation) settings. These settings enable data retrieval enhancements by combining external knowledge with generative AI models. Configure the following options:

Chunk Size: Adjust the size of data chunks for optimal performance based on data volume and memory requirements.

Chunk Overlap: Define overlapping data chunks to ensure continuity for sequential data processing.

Enhanced Website Loader: Enable optimized web-based data ingestion to improve speed and accuracy.