Updated AI docs

implementer-guide/configuration/ai-settings.md

@@ -8,11 +8,64 @@ published: True
 position: 50
 ---
 
-# AI Overview
+<style>
+  img[alt="An image of the providers supported by the Report Server"] {
+  width: 600px;
+}
+
+  img[alt="An image of the Model Name, Server Endpoint, and API Key fields"] {
+  width: 600px;
+}
 
-Beginning with the **2025 Q3** release, we are introducing the new AI Settings to our Report Server. The AI configuration includes two panels — __AI Integration__ and __AI-Powered Report Document Insights__.
+  img[alt="An image of the of the message that will appear if the fields are filled in correctly"] {
+  width: 600px;
+}
 
-During the report preview phase, the [AI-Powered Insights](https://docs.telerik.com/reporting/interactivity/ai-powered-insights) offer an extensive suite of functionalities, including the formulation of responses, the construction of prompts, engagement with AI-generated content, and the execution of predefined instructions.
+  img[alt="An image of the message that will appear if the fields are filled in incorrectly"] {
+  width: 750px;
+}
+
+   img[alt="An image of the Consent Message"] {
+  width: 700px;
+}
+
+   img[alt="Image of the Predefined Prompts in the Report Server"] {
+  width: 600px;
+}
+
+   img[alt="Image of the 'Allow custom prompts' checkbox"] {
+  width: 700px;
+}
+
+   img[alt="An image of the message that will appear if the fields are filled in incorrectly"] {
+  width: 700px;
+}
+
+   img[alt="Image of the message 'The new settings have been saved successfully' in the Report Server Configuration"] {
+  width: 500px;
+}
+
+    img[alt="An Image of how the AI Prompt Consent Dialog Appears in the Report Server"] {
+  width: 900px;
+}
+
+    img[alt="An Image of how the Ask AI Prompt will look with custom questions in the Report Server"] {
+  width: 600px;
+}
+
+    img[alt="An Image of how the Ask AI Prompt will look when the output has been generated in the Report Server"] {
+  width: 600px;
+}
+</style>
+
+
+# AI-Powered Features Settings
+
+With the [**2025 Q3**](https://www.telerik.com/support/whats-new/report-server/release-history/progress-telerik-report-server-2025-q2-11-1-25-716) release, Report Server introduces its first AI-powered feature — a set of configurable AI Settings designed to enhance the reporting experience through intelligent automation. The AI configuration includes two panels — __AI Integration__ and __AI-Powered Report Document Insights__.
+
+During the report preview phase, the [AI-Powered Insights](https://docs.telerik.com/reporting/interactivity/ai-powered-insights) feature provides a rich set of capabilities, including generating responses, building prompts, interacting with AI-generated content, and executing predefined instructions. One of the configurable options in this panel is the ability to display a consent message to end users.
+
+> Report Server does not provide a built-in REST API or include any embedded LLMs/SLMs. To use any AI-powered feature, you must have an active subscription to a third-party LLM provider. This subscription must give you access to a valid REST API endpoint and key, which must be configured in the AI Setup panel.
 
 ![An image of the Report Server with the AI Dialog being opened](../../images/AIPromptReportServer.png)
 
@@ -26,43 +79,51 @@ The AI configuration panel is always visible in the user interface, regardless o
 
 Depending on the state of the `Enable AI` setting:
 
-* If `Enable AI` checkbox is unchecked, the AI configuration section will not be displayed in the Report Server
-* If `Enable AI` checkbox is checked, you will be able to configure the AI by following these steps:
+* If the `Enable AI` checkbox is unchecked, the AI Insights section will not be displayed in the Report Server
+* If the `Enable AI` checkbox is checked, you will be able to configure the AI by following these steps:
 
-1. Choose one of the following providers:
+1. Choose one of the supported providers:
 
-  | Provider |
-  | ------ |
-  |`MicrosoftExtensionsAzureAIInference`|
-  |`MicrosoftExtensionsAzureOpenAI`|
-  |`MicrosoftExtensionsOllama`|
-  |`MicrosoftExtensionsOpenAI`|
+| Provider |
+| ------ |
+|`MicrosoftExtensionsAzureAIInference`|
+|`MicrosoftExtensionsAzureOpenAI`|
+|`MicrosoftExtensionsOllama`|
+|`MicrosoftExtensionsOpenAI`|
 
-  ![An image of the providers supported by the Report Server ](../../images/AIProvidersReportServer.png)
+  ![An image of the providers supported by the Report Server](../../images/AIProvidersReportServer.png)
 
 1. Select a `Model Name`, specify the `Server Endpoint` and the `API Key`
 
   ![An image of the Model Name, Server Endpoint, and API Key fields](../../images/modelNameServiceEndpointAPIKey.png)
 
-> While the **Provider** and **Model** are required for all AI providers, the remaining fields(**Endpoint** and **API Key**) depend on the specific provider's requirements.
+> While the **Provider** and **Model** are required for all AI providers, the remaining fields (**Endpoint** and **API Key**) depend on the specific provider's requirements.
+
+If all required fields are filled in, you can test the integration using the `Test Integration` button. This will send a request to the selected provider with the specified information using a test prompt. If everything is configured correctly, you will receive a response saying `Integration successful`.
 
-If all required fields are filled in, you can test the integration using the `Test Integration` button. This will send a request to the selected provider with the specified information using a test prompt. If everything is configured correctly, you will receive a response saying `Integration successful`. If there is a problem, a pop-up will appear with detailed error information.
+![An image of the of the message that will appear if the fields are filled in correctly](../../images/testIntegrationSuccessfulMessage.png)
 
-![An image of the of the message that will appear if the fields are filled in correctly.](../../images/testIntegrationSuccessfulMessage.png)
+ If there is a problem, a pop-up will appear with detailed error information.
 
-![An image of the message that will appear if the fields are filled in incorrectly.](../../images/testIntegrationFailedMessage.png)
+![An image of the message that will appear if the fields are filled in incorrectly](../../images/testIntegrationFailedMessage.png)
 
 ## AI-Powered Report Document Insights
 
-From this panel, you can choose whether to display a content message by setting the `Show consent message` checkbox to true.
+The **Consent Message Setting** is part of the panel that allows you to configure the AI prompt displayed in the Report Viewer. It plays an important role in ensuring transparency and user awareness when interacting with AI features.
 
-![An image of the Consent Message](../../images/showConsentMessageButton.png)
+From this panel, you can choose whether to display a consent message by setting the `Show consent message` checkbox to true. When this option is activated, a predefined message will be shown to users before they interact with AI-generated content.
 
-> If `Show consent message` checkbox is checked, and the content message field is empty, you will not be able to save the changes.
+The `Consent Message` is a notification shown to users before they interact with AI features. When enabled, the message informs users that by using the AI functionality, they agree to the processing of any data they provide—such as their prompts—for the purpose of delivering the service. It also includes a link to the applicable privacy policy, helping users understand how their data will be handled. You can customize the content of the consent message to better align with your organization’s tone, legal requirements, or user expectations.
+
+![An image of the Consent Message Settings](../../images/showConsentMessage.png)
+
+> If the `Show consent message` checkbox is checked, and the consent message field is empty, you will not be able to save the changes.
 
 ### Prompts Settings
 
-You can create as many **Predefined Prompts** as needed. You can also delete any prompts that are no longer required. However, if there is only one Predefined Prompt, it cannot be deleted. If you would like to enable 'custom prompts' for the end-users and *not* give them any predefined prompts, leave the prompts blank and save.
+The **Prompts Settings** panel allows you to configure how AI prompts are presented to end-users within the Report Viewer. This includes managing predefined prompts, enabling custom prompts, and controlling user interaction with AI-generated content.
+
+You can create as many **Predefined Prompts** as needed. You can also delete any prompts that are no longer required.
 
 ![Image of the Predefined Prompts in the Report Server](../../images/PredefinedPromptsReportServer.png)
 
@@ -91,3 +152,13 @@ After consent is given, the prompt for asking the AI questions will appear in th
 The **Output** of the AI processor will be displayed in the Output tab of the Ask AI Prompt after the result has been generated:
 
 ![An Image of how the Ask AI Prompt will look when the output has been generated in the Report Server](../../images/OutputPromptReportServer.png)
+
+## Differences between Report Server for .NET and Report Server for .NET Framework
+
+In [Report Server for .NET]({%slug coming-soon%}), *Retrieval-Augmented Generation* (RAG) is enabled by default to enhance response accuracy and relevance while minimizing token usage.
+
+Currently, RAG can be disabled only via the `appsettings.json` configuration file. 
+
+Support for controlling the RAG status directly from the UI is planned for upcoming versions of RS.NET. This feature is already logged for Q4.
+
+> Retrieval-Augmented Generation (RAG) is not supported in Report Server for .NET Framework.