Identify RAG Bots as Knowledge Bots
Here we are essentially classifying bots that use Retrieval-Augmented Generation (RAG) as part of the Knowledge Bot category within a platform (like iXHello). This is because their core purpose aligns with what Knowledge Bots are meant to do: provide accurate, context-aware answers based on enterprise knowledge sources.
Why RAG Bots = Knowledge Bots
Knowledge Bots are designed to deliver trusted, data-backed responses.
RAG bots achieve this by:
Retrieving relevant information from enterprise knowledge bases, documents, or APIs.
Generating a natural language response using an LLM, grounded in retrieved content.
This makes them fundamentally different from generic chatbots (which rely only on the model’s memory) and task bots (which execute workflows).
Lets see how we can use RAG Bots as K- Bots
Quick Start App Guide - Studio Mode
Prerequisites
How to use Quick Start App use this document
How to create a KBOT use this
RAG Blueprint Selection
When a user selects a RAG Blueprint in the Quick Start app, an option checkbox will appear with the ability to mark it as a Knowledge Bot.
Knowledge Bot Transformation
By selecting the Knowledge Bot option, the RAG-Bot can be classified as a Knowledge Bot. This change enables the bot to be utilized by employees or other teams using our APIs.

Access the KB contents in Custom App mode by navigating to AI contents.

To make a bot a KBOT, check the KBOT checkbox when creating an app. Alternatively, edit existing apps in the Custom App list.
In Custom Apps, users can upload files under AI Content to provide a knowledge base for the bot.
This means the bot can be configured to retrieve and use enterprise knowledge for answering queries.
Therefore, the option to mark the bot as a Knowledge Bot is shown here, even if the user hasn’t uploaded files yet.
This gives flexibility:
Users can decide upfront that the bot will act as a Knowledge Bot.
Later, they can upload knowledge files to enable that functionality.

Knowledge Bot Configuration in Basic Mode
In Basic Mode apps, the Knowledge Bot checkbox is visible on the Data tab of an existing app. Alternatively, you can create a new app using any blueprint, where the Kbot checkbox will also be available.
This checkbox allows you to mark the app as a knowledge Bot, even if you haven’t uploaded any training files yet.
The reason: Basic Mode supports adding training files later, so the option is kept for flexibility.

Last updated
Was this helpful?