Configure a Retrieval Tool

After creation, you will be navigated to the Retrieval Tool details screen, where you can either keep default settings or continue configuring your tool to tailor it exactly to your needs.

The retrieval Tool details screen is split into the header and three tabs:

Header:

Name
Description
System name. System name is used to identify the Retrieval Tool e.g. when configuring an AI App, so it is strongly recommended not to change it later.

Retrieve tab:

Settings under this tab control how content from defined Knowledge Sources is retrieved and ranked, so that most relevant content is returned by search query.

Magnet AI - Retrieve tab

Knowledge Sources. Connect one or multiple Knowledge Sources that the Retrieval Tool is allowed to search.

Important: it is not recommended to connect Knowledge Sources in multiple languages to one Retrieval Tool.

Retrieval Parameters:

⟩ The similarity score threshold is the minimum score (0 to 1) needed for a content chunk to be retrieved. The higher the value, the stricter the selection criteria. The default is 0.75.

⟩ The number of chunks to select defines how many content chunks matching the similarity score threshold will be returned in search results. For example, if this value is 5, up to 5 top-scored chunks will be returned and displayed as links to the original document. Please note that different chunks can belong to the same original document.

Magnet AI - Retrieve Tools Retrieve tab

Retrieval parameters are closely related to the chunk size set on the Knowledge Source.

⟩ For example, it may be necessary to retrieve additional chunks if they are smaller in size.

Languages Tab:

With settings under this tab, you can track in what languages Magnet AI users communicate with the Retrieval Tool, as well as optimize its performance for multi-lingual use cases.

Magnet AI - Retrieval Tools: Language Tab

Detect question language - if this setting is enabled, the LLM detects the language of user questions. Logs are sent to Azure Application Insights. Language detection can be enabled as an independent feature for monitoring purposes, or as a required part of Multi-lingual Retrieval Tool feature.
Enable multi-lingual Retrieval - this feature significantly improves Retrieval Tool performance for multi-lingual use cases, when users can ask questions in languages other than the Knowledge Source content. To use this feature, the following parameters are necessary:

⟩ Detect question language must be on

⟩ Enable multi-lingual Retrieval must be on

⟩ RAG Tool source language: this is the language of your Knowledge Source content. English is selected by default.

⟩ Translation Prompt Template: Prompt Template that translates user question into the language of content and translates generated answer back into the user’s language. Feel free to use the default prompt template.

UI Settings Tab:

In this tab, you can control the look and feel of your Retrieval Tool

e.g. add a header or help first-time users with some sample questions.

Magnet AI - Retrieval Tools: UI Settings tab

Header configuration - Configure the header and/or subheader that you want to display for your Retrieval Tool. Leave blank to not display any header or subheader.
User feedback - if enabled, users can like or dislike Retrieval search results, and their feedback is logged. This feature is not fully supported yet.
Sample questions - help first-time users formulate their questions by suggesting up to three sample questions.
Show link titles (Upcoming feature) - Control how links to found documents are displayed.
Offer to bypass cache (Upcoming feature) - Gives users an option to bypass semantic cache if caching is enabled and runs a new search across all documents.