About DataMiner Intelligence

Important

DataMiner Intelligence is currently a prerelease feature. Its official release is expected in DataMiner 10.7.1. The functionality is ready to be deployed and tested in production environments, but there may still be significant changes and improvements before the official release. We welcome your feedback at ai@skyline.be.

DataMiner Intelligence covers all agentic AI capabilities of DataMiner. It consists of:

• The DataMiner Intelligence agent framework, with AI agents that help you understand, act, and create on the DataMiner platform.
• Other integrated AI capabilities, such as natural language to GQI and Document Intelligence.

Note

In the context of DataMiner Intelligence, "agents" refers to AI agents. These are not to be confused with DataMiner Agents, i.e., DataMiner nodes. See DataMiner System architecture and components.

DataMiner Intelligence agent framework

The DataMiner Intelligence agent framework is an agentic framework in which AI agents investigate questions, retrieve information, and perform actions in your DataMiner System. The agents can help you:

• Understand: Gain insights into your DataMiner System, for example by investigating alarms, trends, and the state of your elements and services, or by finding answers in the DataMiner documentation.
• Act: Perform controlled operational actions, such as running approved automation scripts to start maintenance or apply a configuration.
• Create: Build and extend solutions on the DataMiner platform.

Agents, context, and tools

The framework comes with out-of-the-box native agents, such as the DataMiner Insights and DataMiner Docs agents.

It is also extensible, so that you can tailor it to your own environment and workflows:

• Add your own context and domain-specific workflows with skills.
• Add your own tools, such as custom data tools to query data and script tools to execute automation scripts.
• Add entire agents of your own with specialized agents, which bundle instructions, skills, and tools into a reusable, purpose-built agent.

Interacting with the agents

You can interact with the agents through the following DataMiner Assistant chat interfaces:

• The DataMiner Assistant app, which is integrated within DataMiner and available from the DataMiner landing page.
• The DataMiner Assistant for Microsoft 365, which gives you access to the agents from Microsoft Teams or Microsoft Copilot.

In addition to chat interaction, it will soon be possible to trigger DataMiner Intelligence from DataMiner Automation. This will allow you to launch agentic workflows from any automated workflow, for example when a new alarm is generated.

Getting started

DataMiner Intelligence requires a connection to dataminer.services and relies on two components that each need to be enabled separately:

• The DataMiner Assistant DxM, which provides the AI capabilities in DataMiner.
• The DataMiner Assistant app, the chat interface that allows you to interact with the agents.

For the steps to enable both components and start exploring DataMiner Intelligence, see Getting started with DataMiner Intelligence.

Other AI capabilities

Next to the agent framework, the DataMiner Assistant DxM also powers the following integrated AI capabilities:

• Natural language to GQI: Describe the data you need in human language, and let DataMiner Assistant create the corresponding GQI query for you in dashboards and low-code apps.
• Document Intelligence: Extract structured information from unstructured documents, such as PDF files, based on instructions you provide, for example as part of an automated workflow.