FAQ Tool in AI Agent
Updated
The FAQ Tool enables your AI Agent to access and deliver answers from training content using Retrieval-Augmented Generation (RAG). This tool is designed to provide accurate, contextual responses to user questions by retrieving relevant knowledge and adapting it into a usable reply.
Steps to Configure the FAQ Tool
Follow the steps below to configure the FAQ Tool in AI Agent Studio.
Access the FAQ Tool
Navigate to AI Agent Studio from the Sprinklr Launchpad.
Select the AI Agent you want to configure.
Expand the Build section and open Tools.
Click + Add Tool and select FAQ Tool from the list.
Define Basic Details
Name: Enter a unique name to identify the tool (for example, Product FAQs or Policy Support FAQs).
Description: Provide a brief description of the tool’s purpose (for example, Delivers answers to product-related queries from the training dataset).
Enter Prompt
In the Prompt field, specify the input that guides how the FAQ content should be retrieved.
Example: “Retrieve the most relevant answer for the user’s query about account setup.”
Prompts should be concise, goal-oriented, and aligned with the intended FAQ source.
Add Additional Field
Use the Additional Field option to enter or select values, or link to a specific resource (such as a dataset, FAQ knowledge base, or category filter).
This helps refine results and ensures the FAQ tool retrieves answers from the correct source.
Adapt Response
Use the Adapt Response option to enable customization of how the retrieved answer is presented.
Adapted Response Variable: Enter the variable name where the adapted response will be stored.
Example: Store the adapted response in a variable like faqResponse for reuse in downstream tasks or workflows.
Save Configuration
Once all details are entered, click Save.
The FAQ Tool will now be available in the Tools Record Manager and can be integrated into your AI Agent workflows.
Best Practices
Keep prompts specific to reduce ambiguity in FAQ retrieval.
Use Additional Fields to narrow down search scope (e.g., filter answers by category such as Billing or Technical Support).
Always configure an Adapted Response Variable so the AI Agent can reuse or display the retrieved content consistently across different conversation flows.