Nerdio Copilot leverages an AI-based assistant to quickly search for information about Nerdio Manager, its features, and functions.
Note
Availability of Azure OpenAI services is limited and varies by Azure region.
Important
Nerdio Copilot has been completely rebuilt and refactored in NME 8.0. If you have previously deployed Nerdio Copilot, in NME 7.7 or previous versions, you will need to disable the legacy version Nerdio Copilot.
Note
In Nerdio Manager 8.3, the Nerdio Manager Copilot and Azure AI settings are merged, so you configure AI services in one place. Description generation and auto-scale analytics recommendations are now Nerdio Manager Copilot settings. The Azure AI area under System > Settings > Integrations no longer has an OpenAI configuration. Only the PII Detector remains there. Azure AI features can use Foundry resources, and Nerdio Manager Copilot can use an existing Foundry resource without a separate deployment.
To uninstall the legacy version of Nerdio Copilot:
-
In Nerdio Manager, navigate to System > Settings > Nerdio.
-
Scroll down and expand the Nerdio Manager Copilot section.
-
Select Configure.
This will show all the resources that supported the legacy version of Nerdio Copilot.
-
Select OK to remove these resources.
To enable Nerdio Copilot:
-
In Nerdio Manager, navigate to System > Settings > Nerdio.
-
Scroll down and expand the Nerdio Manager Copilot section.
-
Select Enabled. The Deploy Nerdio Manager Copilot dialog box opens.
-
Under Foundry Configuration, enter the following information:
-
Foundry: From the drop-down list, select an existing Foundry resource. To create a new Foundry resource, type a name for it, and then select Create followed by the name that you typed.
-
Foundry deployment: Select an existing Foundry deployment, or create a new deployment with a custom name. You can also create a new deployment in an existing Foundry resource.
-
Resource group: From the drop-down list, select the resource group to which you want to assign all the resources required to run Nerdio Manager Copilot.
-
Model: From the drop-down lists, select the OpenAI model based on available quotas.
-
Deployment type: Choose deployment type. Either Data Zone Standard or Global Standard.
-
-
Under Nerdio Manager Permissions, use the Read toggle to control Nerdio Manager Copilot's read access to Nerdio Manager data.
-
Under Nerdio Manager Copilot Settings, enter the following information:
-
Description generation: Toggle On to let users generate descriptions for host pools and Unified Application Management deployment policies with Nerdio Manager Copilot. See Overview of AI-Powered Description Generation.
-
Auto-scale analytics recommendations: Toggle On to enable auto-scale insights recommendations. See Auto-scale Insights - Intelligent Recommendations.
-
Recommendations preloading: Select this option to have the auto-scale insights recommendations automatically preloaded. Otherwise, the recommendations are not loaded until a user requests them.
-
-
Optionally, expand Customize resources tags, and then enter the Azure tags to apply to the Nerdio Manager Copilot resources.
-
Once you have entered all the desired information, select OK.
The deployment task starts and takes about 2-5 minutes depending on the Azure region and other conditions. You can follow the task's progress in the Tasks section.
Once you've deployed all the required Azure resources, you can use Nerdio Copilot to help navigate Nerdio Manager.
Note
Nerdio Manager Copilot caches the Nerdio Help Center KBs to answer your questions and refreshes the cache on a daily basis.
To use Nerdio Copilot:
-
In the upper-right corner of Nerdio Manager, select the Nerdio Manager Copilot
icon.
Nerdio Manager connects to the bot and downloads the chat history.
-
Ask any question related to Nerdio Manager and send them to Copilot. After a short period of time, Copilot will display an answer.
-
Optionally, select Citations to view a list of cited Knowledge Base articles.
-
You may select either the sidebar window or detached window view options.
You may adjust Nerdio Copilot's settings at any time.
To manage Nerdio Copilot's chat settings:
-
Select the Settings
icon.
-
Enter the following information:
-
Intent recognition service OpenAI Model: From the drop-down list, select the desired model.
-
Max days to retain history: Enter the number of chat histories to retain.
-
-
Once you have entered the desired information, select Save.
-
When prompted, select Confirm to confirm your changes.
Copilot uses Azure OpenAI, powered by Large Language Model (LLM) that has been augmented with Nerdio-specific information. Due to this being an LLM, answers are not deterministic. For more details, see Introduction to large language models.
When you notice an incorrect answer, you can submit feedback in the following ways:
-
Enter a comment (optional).
-
Follow Nerdio’s standard support escalation path.
Follow this procedure to disable Nerdio Copilot.
Note
When you disable Nerdio Copilot, you can keep the Azure resources that were deployed when the feature was enabled, or clean them up. If you clean them up, all of the resources are removed except for the Smart detector alert rule.
To disable Nerdio Copilot:
-
In Nerdio Manager, navigate to System > Settings > Nerdio.
-
Scroll down and expand the Nerdio Manager Copilot section, and select Configure.
-
When prompted, choose whether to keep the resources or clean them up, and then select OK.
The disable task starts and takes about 2-3 minutes depending on the Azure region and other conditions. You can follow the task's progress in the Tasks section.
Nerdio Copilot includes Script Pro, which allows you to build scripts with AI assistance.
Warning
Ensure that all AI-generated scripts are tested and validated manually before deployment to your production environments.
To build a script with Nerdio Copilot:
-
Open Nerdio Copilot.
-
Ask a question related to building a script. For example:
Can you help me generate a management script for a Windows VM to set this regkey? [HKEY_LOCAL_MACHINE\SOFTWARE\MyApp]"NewFeatureEnable"=dword:1
-
Follow the prompts and reply as needed.
-
When the script is generated, copy the code.
AnalyticsPro, is a powerful tool within Nerdio Copilot that transforms your natural language questions into Kusto Query Language (KQL) queries. With AnalyticsPro, you can effortlessly generate KQL queries that can be executed in your Azure Log Analytics workspace, unlocking valuable insights from your data without needing to master complex query syntax.
To generate a KQL query using AmalyticsPro:
-
Open Nerdio Copilot.
-
Enter the details of the KQL query you want to generate. For example:
Can you write me a KQL query to get the number of active users from the past hour?
-
Select the send
icon.
-
The KQL query is generated.
-
You can interact with AnalyticsPro and ask for additional information based on the results, or provide options to help refine your query. Respond to these prompts as needed to tailor the query to your specific requirements.
-
Once complete, you can copy the query and run it in your Azure Log Analytics workspace.
Warning
Ensure that all AI-generated queries are tested and validated manually before deployment to your production environments
The cost of the feature depends on its usage. The cost is per tokens used, with costs varying depending on which model is chosen and how much it is used.
If the feature is enabled but not used, there is no standing cost.
LLM implementation work with tokens. The number of tokens is a combination of system prompt, input from user, and output from LLM. The number of tokens defines how much “memory” about a previous exchange in the current conversation the bot has. If a conversation is long, larger than the max token count configured for the model, older data is dropped. However, we expose all the chat history in the UI until the chat history is deleted using the delete history button.
By default, Nerdio Copilot deploys all resources in the same region. Azure OpenAI resources can be created in different regions, based on user’s selection. When possible and applicable, we deploy the free or the lowest paid tier resources; this is not configurable.
Comments (0 comments)