Creating data sources in the Pega GenAI Knowledge Buddy

This demo will show you how to create data sources in Pega GenAI Knowledge Buddy™.

Myco, a telecommunications company, utilizes Pega Knowledge Buddy to provide quick, accurate, and consistent answers to its customer questions.

The company's initial objective is to create a data source and push textual-based content into it for one of its trusted customers, Reliance Industries Limited. This enables Knowledge Buddy to use the data source for providing personalized answers to customer queries. Additionally, the company aims to grant viewing access to all its users, for easy retrieval of information from the data source.

To implement the business scenario, in the navigation pane of the Pega Knowledge Buddy Portal, click Data Source. At the top right of the portal, click Create data source.

Individuals with the access role of KnowledgeBuddy:Admin or KnowledgeBuddy:DataSourceManager have the authority to create new data sources. If users do not have these roles, they cannot create a data source in the application.

In the Name field of the Create Data Source dialog box, enter some text to identify the data source. For example, CustomerRelianceIn.

In the Collection list, select RelianceIndustriesLTD. This collection was created for the specific purpose of storing data sources connected to this customer.

In the Description field, enter some text to describe this data source. For example, A data source for the customer, Reliance Industries Limited.

By default, when creating a data source, the Data source manager is given Manage data source access, and the Knowledge buddy manager role is given with View data source access. Users can modify access roles with certain access types from the Quick Create view of the data source by adding or deleting access roles.

In the Access configurations section, click Add to add one or more access types and roles.

In the Type and Role name lists, select a task that can be performed on the data source and the role name that must be applied to individuals so that they can perform the task.

In this scenario, the company wants to grant all their users the opportunity to view data sources. To do that, select the View data source access type and the Knowledge buddy public role from the Type and Role drop-down lists.

Click Advanced settings to set content processing options for the data source, such as the Preferred extraction method or the Chunking method.

For the extraction method, you can choose between the Standard extraction method or the GenAI extraction method.

The Standard extraction method is the default method used for plain text or unsupported file formats. It extracts basic text, including text within tables, but does not extract non-text elements such as images or table structures.

Choose the GenAI extraction method for supported file types to retain tabular formats and other structured elements. It ensures the integrity of structured data during extraction. If the GenAI extraction process is unsuccessful, the system automatically reverts to the Standard method.

By default, you set the Chunking method at a Data collection level, but you can choose a different method at Data source level. This overrides the setting you chose at a Data collection level for this Data source only.

Click Submit.

After you create the data source, create and ingest some content then incorporate the data source into a Buddy. This enables the Buddy to utilize the ingested content when composing responses to user queries.

If you are creating content types in the Pega Knowledge portal, then those content types create a corresponding data source for Pega Knowledge Buddy with the name Knowledge_ {ContentType ID}. For example, if there is a content type with the name Diagnostics, then the data source corresponding to this would be Knowledge_Diagnostics.

You have reached the end of this demo. You have learned:

• How to create data sources in the Pega Knowledge Buddy.