Merge branch 'main' of https://github.com/MicrosoftDocs/azure-ai-docs-pr into foundry-move

Author: aahill

articles/ai-foundry/agents/concepts/standard-agent-setup.md

@@ -27,18 +27,17 @@ Both standard setup configurations are designed to give you complete control ove
 
 By bundling these BYO features (file storage, search, and thread storage), the standard setup guarantees that your deployment is secure by default. All data processed by Azure AI Foundry Agent Service is automatically stored at rest in your own Azure resources, helping you meet internal policies, compliance requirements, and enterprise security standards. 
 
-## Project-Level Data Isolation
+### Azure Cosmos DB for NoSQL
+
+Your existing Azure Cosmos DB for NoSQL Account used in standard setup must have a total throughput limit of at least **3000 RU/s**. Both **Provisioned Throughput** and **Serverless** modes are supported.
 
-Azure AI Foundry enforces project-level data isolation by default. When you configure your own resources in the project capability host: 
-* **Azure Storage**: Two Blob containers are automatically provisioned: 
-    * One for uploaded files 
-    * One for intermediate system data (for example, chunks, embeddings) 
-* **Azure Cosmos DB**: Three containers are provisioned under a dedicated enterprise_memory database: 
-    * thread-message-store: End-user conversations 
-    * system-thread-message-store: Internal system messages 
-    * agent-entity-store: Model inputs and outputs 
+When you use standard setup, **three containers** will be provisioned in your existing Cosmos DB account, and **each container requires 1000 RU/s**.
+* thread-message-store: End-user conversations
+* system-thread-message-store: Internal system messages
+* agent-entity-store: Agent metadata including their instructions, tools, name, etc.
 
-This default behavior was chosen to reduce configuration complexity while still enforcing strict data boundaries—ensuring each project has a clean, isolated storage footprint without requiring manual setup. 
+## Project-Level Data Isolation
+Standard setup enforces project-level data isolation by default. Two blob storage containers will automatically be provisioned in your storage account, one for files and one for intermediate system data (chunks, embeddings) and three containers will be provisioned in your Cosmos DB, one for user systems, one for system messages, and one for user inputs related to created agents such as their instructions, tools, name, etc. This default behavior was chosen to reduce setup complexity while still enforcing strict data boundaries between projects.
 
 ## Capability hosts
 **Capability hosts** are sub-resources on both the Account and Project, enabling interaction with the Azure AI Foundry Agent Service. 
@@ -85,4 +84,4 @@ This default behavior was chosen to reduce configuration complexity while still
         * Assign role: Cosmos DB Built-in Data Contributor 
     * Cosmos DB for NoSQL container: `<'${projectWorkspaceId}>-agent-entity-store'` 
         * Assign role: Cosmos DB Built-in Data Contributor 
-11. Once all resources are provisioned, all developers who want to create/edit agents in the project should be assigned the role: Azure AI User on the project scope. 
+11. Once all resources are provisioned, all developers who want to create/edit agents in the project should be assigned the role: Azure AI User on the project scope.

articles/ai-foundry/agents/environment-setup.md

@@ -18,14 +18,14 @@ Creating your first agent with Azure AI Foundry Agent Service is a two-step proc
 1. Set up your agent environment.
 1. Create and configure your agent using either the SDK of your choice or the Azure Foundry Portal. 
 
-Use this article to learn more about setting up your agents.
+Use this article to learn more about setting up your agent environment.
 
 ### Required permissions 
 
 | Action                                                                 | Required Role                   |
 |------------------------------------------------------------------------|----------------------------------|
 | Create an account and project                                          | Azure AI Account Owner           |
-| **Standard Setup Only:** Assign RBAC for required resources (Cosmos DB, Search, Storage, etc.) | Role Based Access Administrator  |
+| **Standard Setup Only:** Assign RBAC for required resources (Cosmos DB, Search, Storage, etc.) | Role Based Access Control Administrator  |
 | Create and edit agents                                                 | Azure AI User                    |
 
 ## Set up your agent environment
@@ -37,7 +37,7 @@ Agents are scoped at the project level, which ensures data isolation—agents wi
 * An Azure subscription - [Create one for free](https://azure.microsoft.com/free/cognitive-services).
 * Ensure that the individual creating the account and project has the **Azure AI Account Owner** role at the subscription scope
 * If configuring **Standard Setup**, the same individual must also have permissions to assign roles to required resources (Cosmos DB, Search, Storage).
-    * The built-in role needed is **Role Based Access Administrator**.
+    * The built-in role needed is **Role Based Access Control Administrator**.
     * Alternatively, having the **Owner** role at the subscription level also satisfies this requirement.
     * The key permission needed is: `Microsoft.Authorization/roleAssignments/write`
 

articles/ai-foundry/agents/how-to/use-your-own-resources.md

@@ -85,6 +85,9 @@ Use an existing AI Services / Azure OpenAI, Azure Storage account, Azure Cosmos
     ```
 
 ### Use an existing Azure Cosmos DB for NoSQL account for thread storage
+**Azure Cosmos DB for NoSQL**
+- Your existing Azure Cosmos DB for NoSQL Account used in standard setup must have at least a total throughput limit of at least 3000 RU/s. Both Provisioned Thoughtput and Serverless are supported.
+    - 3 containers will be provisioned in your existing Cosmos DB account and each need 1000 RU/s
 
 1. To get your Azure Cosmos DB account resource ID, sign in to the Azure CLI and select the subscription with your account: 
     

articles/ai-foundry/concepts/architecture.md

@@ -16,8 +16,7 @@ author: Blackmist
 
 # Azure AI Foundry architecture 
 
-> [!NOTE]
-> The architecture discussed in this article is specific to a **[!INCLUDE [hub](../includes/hub-project-name.md)]**. For more information, see [Types of projects](../what-is-azure-ai-foundry.md#project-types).
+[!INCLUDE [hub-only-alt](../includes/uses-hub-only-alt.md)]
 
 Azure AI Foundry provides a unified experience for AI developers and data scientists to build, evaluate, and deploy AI models through a web portal, SDK, or CLI. Azure AI Foundry is built on capabilities and services provided by other Azure services.
 

articles/ai-foundry/concepts/encryption-keys-portal.md

@@ -117,6 +117,9 @@ Managed identity must be enabled as a prerequisite for using customer-managed ke
 
 Customer-managed key encryption is configured via Azure portal in a similar way for each Azure resource:
 
+> [!IMPORTANT]
+> The Azure Key Vault used for encryption **must be in the same resource group** as the AI Foundry project. Key Vaults in other resource groups are not currently supported by the deployment wizards or project configuration workflows.
+
 1. Create a new Azure AI Foundry resource in the [Azure portal](https://portal.azure.com/).
 1. Under the **Encryption** tab, select **Customer-managed key**, **Select vault and key**, and then select the key vault and key to use.
 

articles/ai-foundry/concepts/evaluation-evaluators/agent-evaluators.md

@@ -70,7 +70,7 @@ intent_resolution(
 
 ### Intent resolution output
 
-The numerical score on a likert scale (integer 1 to 5) and a higher score is better. Given a numerical threshold (default to 3), we also output "pass" if the score <= threshold, or "fail" otherwise. Using the reason and additional fields can help you understand why the score is high or low.
+The numerical score on a likert scale (integer 1 to 5) and a higher score is better. Given a numerical threshold (default to 3), we also output "pass" if the score >= threshold, or "fail" otherwise. Using the reason and additional fields can help you understand why the score is high or low.
 
 ```python
 {
@@ -137,7 +137,7 @@ tool_call_accuracy(
 
 ### Tool call accuracy output
 
-The numerical score (passing rate of correct tool calls) is 0-1 and a higher score is better. Given a numerical threshold (default to 3), we also output "pass" if the score <= threshold, or "fail" otherwise. Using the reason and tool call detail fields can help you understand why the score is high or low.
+The numerical score (passing rate of correct tool calls) is 0-1 and a higher score is better. Given a numerical threshold (default to 3), we also output "pass" if the score >= threshold, or "fail" otherwise. Using the reason and tool call detail fields can help you understand why the score is high or low.
 
 ```python
 {
@@ -174,7 +174,7 @@ task_adherence(
 
 ### Task adherence output
 
-The numerical score on a likert scale (integer 1 to 5) and a higher score is better. Given a numerical threshold (default to 3), we also output "pass" if the score <= threshold, or "fail" otherwise. Using the reason field can help you understand why the score is high or low.
+The numerical score on a likert scale (integer 1 to 5) and a higher score is better. Given a numerical threshold (default to 3), we also output "pass" if the score >= threshold, or "fail" otherwise. Using the reason field can help you understand why the score is high or low.
 
 ```python
 {

articles/ai-foundry/concepts/evaluation-evaluators/general-purpose-evaluators.md

@@ -59,7 +59,7 @@ coherence(
 
 ### Coherence output
 
-The numerical score on a likert scale (integer 1 to 5) and a higher score is better. Given a numerical threshold (default to 3), we also output "pass" if the score <= threshold, or "fail" otherwise. Using the reason field can help you understand why the score is high or low.
+The numerical score on a likert scale (integer 1 to 5) and a higher score is better. Given a numerical threshold (default to 3), we also output "pass" if the score >= threshold, or "fail" otherwise. Using the reason field can help you understand why the score is high or low.
 
 ```python
 {
@@ -88,7 +88,7 @@ fluency(
 
 ### Fluency output
 
-The numerical score on a likert scale (integer 1 to 5) and a higher score is better. Given a numerical threshold (default to 3), we also output "pass" if the score <= threshold, or "fail" otherwise. Using the reason field can help you understand why the score is high or low.
+The numerical score on a likert scale (integer 1 to 5) and a higher score is better. Given a numerical threshold (default to 3), we also output "pass" if the score >= threshold, or "fail" otherwise. Using the reason field can help you understand why the score is high or low.
 
 ```python
 {
@@ -127,7 +127,7 @@ qa_eval(
 
 ### QA output
 
-While F1 score outputs a numerical score on 0-1 float scale, the other evaluators output numerical scores on a likert scale (integer 1 to 5) and a higher score is better. Given a numerical threshold (default to 3), we also output "pass" if the score <= threshold, or "fail" otherwise. Using the reason field can help you understand why the score is high or low.
+While F1 score outputs a numerical score on 0-1 float scale, the other evaluators output numerical scores on a likert scale (integer 1 to 5) and a higher score is better. Given a numerical threshold (default to 3), we also output "pass" if the score >= threshold, or "fail" otherwise. Using the reason field can help you understand why the score is high or low.
 
 ```python
 {

articles/ai-foundry/concepts/evaluation-evaluators/rag-evaluators.md

@@ -63,7 +63,7 @@ retrieval(
 
 ### Retrieval output
 
-The numerical score on a likert scale (integer 1 to 5) and a higher score is better. Given a numerical threshold (a default is set), we also output "pass" if the score <= threshold, or "fail" otherwise. Using the reason field can help you understand why the score is high or low.
+The numerical score on a likert scale (integer 1 to 5) and a higher score is better. Given a numerical threshold (a default is set), we also output "pass" if the score >= threshold, or "fail" otherwise. Using the reason field can help you understand why the score is high or low.
 
 ```python
 {
@@ -163,7 +163,7 @@ document_retrieval_evaluator(retrieval_ground_truth=retrieval_ground_truth, retr
 
 ### Document retrieval output
 
-All numerical scores have `high_is_better=True` except for `holes` and `holes_ratio` which have `high_is_better=False`. Given a numerical threshold (default to 3), we also output "pass" if the score <= threshold, or "fail" otherwise. 
+All numerical scores have `high_is_better=True` except for `holes` and `holes_ratio` which have `high_is_better=False`. Given a numerical threshold (default to 3), we also output "pass" if the score >= threshold, or "fail" otherwise. 
 
 ```python
 {
@@ -206,7 +206,7 @@ groundedness(
 
 ### Groundedness output
 
-The numerical score on a likert scale (integer 1 to 5) and a higher score is better. Given a numerical threshold (default to 3), we also output "pass" if the score <= threshold, or "fail" otherwise. Using the reason field can help you understand why the score is high or low.
+The numerical score on a likert scale (integer 1 to 5) and a higher score is better. Given a numerical threshold (default to 3), we also output "pass" if the score >= threshold, or "fail" otherwise. Using the reason field can help you understand why the score is high or low.
 
 ```python
 {
@@ -276,7 +276,7 @@ relevance(
 
 ### Relevance output
 
-The numerical score on a likert scale (integer 1 to 5) and a higher score is better. Given a numerical threshold (default to 3), we also output "pass" if the score <= threshold, or "fail" otherwise. Using the reason field can help you understand why the score is high or low.
+The numerical score on a likert scale (integer 1 to 5) and a higher score is better. Given a numerical threshold (default to 3), we also output "pass" if the score >= threshold, or "fail" otherwise. Using the reason field can help you understand why the score is high or low.
 
 ```python
 {
@@ -306,7 +306,7 @@ response_completeness(
 
 ### Response completeness output
 
-The numerical score on a likert scale (integer 1 to 5) and a higher score is better. Given a numerical threshold (default to 3), we also output "pass" if the score <= threshold, or "fail" otherwise. Using the reason field can help you understand why the score is high or low.
+The numerical score on a likert scale (integer 1 to 5) and a higher score is better. Given a numerical threshold (default to 3), we also output "pass" if the score >= threshold, or "fail" otherwise. Using the reason field can help you understand why the score is high or low.
 
 ```python
 {

articles/ai-foundry/concepts/evaluation-evaluators/textual-similarity-evaluators.md

@@ -58,7 +58,7 @@ similarity(
 
 ### Similarity output
 
-The numerical score on a likert scale (integer 1 to 5) and a higher score means a higher degree of similarity. Given a numerical threshold (default to 3), we also output "pass" if the score <= threshold, or "fail" otherwise. Using the reason field can help you understand why the score is high or low.
+The numerical score on a likert scale (integer 1 to 5) and a higher score means a higher degree of similarity. Given a numerical threshold (default to 3), we also output "pass" if the score >= threshold, or "fail" otherwise. Using the reason field can help you understand why the score is high or low.
 
 ```python
 {
@@ -87,7 +87,7 @@ f1_score(
 
 ### F1 score output
 
-The numerical score is a 0-1 float and a higher score is better. Given a numerical threshold (default to 0.5), we also output "pass" if the score <= threshold, or "fail" otherwise.
+The numerical score is a 0-1 float and a higher score is better. Given a numerical threshold (default to 0.5), we also output "pass" if the score >= threshold, or "fail" otherwise.
 
 ```python
 {
@@ -115,7 +115,7 @@ bleu_score(
 
 ### BLEU output
 
-The numerical score is a 0-1 float and a higher score is better. Given a numerical threshold (default to 0.5), we also output "pass" if the score <= threshold, or "fail" otherwise.
+The numerical score is a 0-1 float and a higher score is better. Given a numerical threshold (default to 0.5), we also output "pass" if the score >= threshold, or "fail" otherwise.
 
 ```python
 {
@@ -144,7 +144,7 @@ gleu_score(
 
 ### GLEU score output
 
-The numerical score is a 0-1 float and a higher score is better. Given a numerical threshold (default to 0.5), we also output "pass" if the score <= threshold, or "fail" otherwise.
+The numerical score is a 0-1 float and a higher score is better. Given a numerical threshold (default to 0.5), we also output "pass" if the score >= threshold, or "fail" otherwise.
 
 ```python
 {
@@ -173,7 +173,7 @@ rouge(
 
 ### ROUGE score output
 
-The numerical score is a 0-1 float and a higher score is better. Given a numerical threshold (default to 0.5), we also output "pass" if the score <= threshold, or "fail" otherwise.
+The numerical score is a 0-1 float and a higher score is better. Given a numerical threshold (default to 0.5), we also output "pass" if the score >= threshold, or "fail" otherwise.
 
 ```python
 {
@@ -208,7 +208,7 @@ meteor_score(
 
 ### METEOR score output
 
-The numerical score is a 0-1 float and a higher score is better. Given a numerical threshold (default to 0.5), we also output "pass" if the score <= threshold, or "fail" otherwise.
+The numerical score is a 0-1 float and a higher score is better. Given a numerical threshold (default to 0.5), we also output "pass" if the score >= threshold, or "fail" otherwise.
 
 ```python
 {

articles/ai-foundry/concepts/fine-tuning-overview.md

@@ -29,8 +29,6 @@ Consider fine-tuning GenAI models to:
 
 [Azure AI Foundry](https://ai.azure.com/?cid=learnDocs) offers several models across model providers enabling you to get access to the latest and greatest in the market. [View this list for more details](#supported-models-for-fine-tuning). 
 
-:::image type="content" source="../media/concepts/model-catalog-fine-tuning.png" alt-text="Screenshot of Azure AI Foundry model catalog and filtering by Fine-tuning tasks." lightbox="../media/concepts/model-catalog-fine-tuning.png":::
-
 ## Getting started with fine-tuning
 
 When starting out on your generative AI journey, we recommend you begin with prompt engineering and RAG to familiarize yourself with base models and its capabilities.