AI Capabilities

LeanIX AI capabilities use generative AI to streamline enterprise architecture tasks, including querying data and generating documentation. They enhance efficiency and understanding, empowering users with actionable insights.

Overview

LeanIX AI capabilities apply the power of generative AI to enterprise architecture data. By automating tedious documentation tasks, accelerating data usage, and speeding up research, LeanIX AI capabilities unlock tangible benefits that help businesses gain a competitive edge.

LeanIX AI capabilities allow admins to explore the power of large language models flexibly and securely. They are powered by dedicated OpenAI models hosted on Microsoft Azure, ensuring the security and confidentiality of enterprise architecture data.

📘

Note

LeanIX AI capabilities leverage, Microsoft’s Azure OpenAI Service with Azure’s security and enterprise promise. The data from LeanIX workspaces is NOT used to train large language models. For further details on data privacy and security, see the Azure documentation.

Base AI Capabilities

Base AI capabilities in LeanIX encompass the following features:

  • Inventory AI prompt leverages generative AI to provide a natural language interface for querying data, generating documentation, and obtaining insights in the inventory.
  • AI-assisted text simplifies and standardizes the process of creating fact sheet descriptions by analyzing content and context. Additionally, it transforms and enriches existing text to ensure clarity and accuracy.
  • AI-generated context enriches to-dos with additional details and explanations, aiding users in understanding tasks better and completing them efficiently.
  • AI-supported translation enables you to translate labels and help texts for fact sheet attributes in the meta model configuration.

Activating Base AI Capabilities

The activation process includes two steps:

  1. Requesting the feature and signing off AI terms for your organization.
  2. Activating the AI capabilities for each workspace separately, in case your organization has multiple ones.

📘

  • Signing off the AI terms is a requirement per contract, while activating the base AI capabilities is done individually for each workspace.
  • LeanIX base AI capabilities doesn't incur any additional costs.

Admins can request and then activate the base AI capabilities by following these steps:

  1. Navigate to Admin settings and select Optional Features & Early Access.

  2. Click Request Access on the base AI capabilities feature.

    Requesting Access for Base AI Capabilities

    Requesting Access for Base AI Capabilities

  3. You are prompted to agree to the AI terms. Complete the process by clicking on Open the AI Terms and providing your consent.

  4. Once you have successfully agreed to the terms, click Activate on the base AI capabilities feature in Optional Features & Early Access.

📘

For further information on AI terms, user data handling, Microsoft Azure OpenAI usage, and privacy concerns, refer to Frequently Asked Questions.

When you no longer need the AI capabilities, you can deactivate them by selecting Deactivate next to Base AI Capabilities in the Optional Features & Early Access section.

Inventory AI Prompt

Inventory AI prompt combines the information in your inventory with the power of generative AI to make enterprise architecture information more accessible and actionable. It provides a natural language interface for interacting with your enterprise architecture repository, allowing you to ask questions, query data, and receive insights using everyday language. You can use predefined requests or your own prompts to get assistance in various tasks.

📘

Note

The inventory AI prompt feature is available only for admins.

With the inventory AI prompt, you can:

  • Generate descriptions and get better context.
  • Request recommendations for optimizing enterprise architecture based on best practices, industry standards, and organizational goals.
  • Query for suggestions on changes to application portfolios, infrastructure configurations, and IT governance processes.
  • Obtain details about specific data objects, such as their attributes, dependencies, and usage, and get assistance in data object-related operations through REST API examples.
  • Obtain quick insights regarding potential risks, relevant regulations, and standards.

Using the Inventory AI Prompt

To use the inventory AI prompt, do the following:

  1. Click Start AI prompt in the inventory to open the Inventory AI prompt page.
  2. Enter your prompt in the prompt field, or select predefined prompts from the drop-down menu. Predefined prompts also automatically apply filters as required.
  3. Click on the Run Request button (play icon) next to the fact sheets for which you want to execute the request. You can also select multiple fact sheets and run the request at once for selected fact sheets.

    📘

    The selection of all fact sheets in scope for bulk action is intentionally restricted to prevent unnecessary load and resource usage on our end.

  4. To ensure accuracy, it's recommended to review AI-generated content before saving it to the fact sheet. Therefore, you need to confirm each response individually by clicking the Save icon represented by the check mark.
  5. If the response is not satisfactory, you can regenerate it by clicking the Run Request button (play icon).
Using Inventory AI Prompt

Using Inventory AI Prompt

👍

Tip

  • Adjust the predefined prompt or create your own request to make it more specific to your needs and to improve the results.
  • By specifying the language in your requests, you can receive responses in different languages. For example, "Create description of 2 sentences in French with link to French website".

AI-Assisted Text

The AI-assisted text allows you to effortlessly create or enhance various types of text content by analyzing the context and providing options to improve it through rewriting or summarizing existing text.

For example, currently, AI-assisted text analyzes the content and context of the fact sheet to generate a concise and accurate description for fact sheets. This feature streamlines the process of creating informative descriptions, saving users time and effort.

To use the AI-assisted text feature, click on the AI help icon located at the bottom right corner of the description field. It automatically generates a description, which you can then refine, rewrite, summarize, or format as needed. Additionally, you can explain the relations of all related fact sheets, create bullet points, format them as paragraphs, or undo changes.

AI-assisted Text in Description of a Fact Sheet

AI-Assisted Text in Description of a Fact Sheet

AI-Generated Context

AI-generated context provides additional information and insights by analyzing existing data and context, helping users better understand tasks or requests and make informed decisions.

For example, currently, the AI-generated context provides additional details and explanations for the to-dos to help you understand the task better and complete it more efficiently. It recognizes the underlying request of the to-do and considers the to-do’s description and other details in the fact sheet to generate context and help users in their tasks.

To use the AI-generated context feature in to-dos, click on the AI Help icon in the task. You can regenerate the response if needed and delete the context when no longer required.

AI-generated Context in To-Dos

AI-Generated Context in To-Dos

AI-Supported Translation

The AI-supported translation allows administrators to add and translate labels and help texts for both newly created and existing attributes in the meta model configuration. This enables you to ensure that labels and help texts for fact sheet attributes are accessible in multiple languages, facilitating a more inclusive and comprehensible user experience for all users, regardless of language preference. The ability to perform translations en masse saves time and effort while also ensuring consistency across the translated content.

To use AI-supported translation while adding a new attribute, click the Add Translations button on the Translations tab on the right-side panel. Labels and help texts in different languages are automatically generated for all the fields, field values, and relations.

Adding Translations to Attributes Using AI-Supported Translation

Adding Translations to Attributes Using AI-Supported Translation

If needed, you can undo or rewrite the generated translations. For existing attributes, if a translation is missing in one or more languages, you can add the missing translations by clicking the AI Help icon and selecting the appropriate option.

Options for Adding Missing Translations, Rewriting, or Undoing the Translation

Options for Adding Missing Translations, Rewriting, or Undoing the Translation

Frequently Asked Questions

Usage and General Queries

Q: If I have multiple workspaces, which of them gets the feature?

A: All LeanIX Enterprise Architecture workspaces under your main LeanIX contract receive the feature. However, the feature has to be activated for each workspace individually. See Activating Base AI Capabilities.

Q: Do LeanIX AI capabilities use a dedicated language model for each LeanIX customer?

A: No, LeanIX AI capabilities leverage Microsoft's Azure OpenAI Service with the same unmodified model(s) for all LeanIX customers.

Q: Can I use a different language model or my own Azure OpenAI Service?

A: No, the LeanIX AI capabilities only leverage the dedicated OpenAI models of Microsoft's Azure OpenAI Service.

Q: Where can I find more information on the used AI Service?

A: You can find more information on the website of Microsoft’s Azure OpenAI Service.

Q: Are there API endpoints available to the base AI capabilities?

A: No, the base AI capabilities cannot be accessed via our API endpoint

Data Handling and Privacy Concerns

Q: Where is the language model hosted, and where is the data processed?

A: Depending on the availability, we try to leverage the language model hosted in the customer's region. However, as Microsoft does not guarantee availability in all regions, we might choose different regions where necessary.

Q: Which data is processed by the Microsoft Azure OpenAI service?

A: The data processed depends on the specific use case. For instance, to generate a description for a Fact Sheet, relevant data from that fact sheet, such as name, existing description, and relations are used. When generating the context of a to-do, relevant data from any related fact sheet is also taken into account.

It's important to note that no personally identifiable information (PII) data, like names or email addresses of users, is sent to the Azure OpenAI service.

Q: Which data from my LeanIX AI capabilities usage does LeanIX store?

A: The request query in the prompt and the selected fact sheet type info are stored to analyze usage patterns. The query context with workspace data is not stored.

Q: Does anyone outside of LeanIX have access to the processed data?

A: No. No explicit access is granted to any third party, including Microsoft Azure personnel.

Q: Is any of my workspace data used to train the AI/language model?

A: No, the large language models used are not retrained with data from your workspace.

Q: Does LeanIX or Microsoft Azure OpenAI send my workspace data to OpenAI?

A: No. Microsoft hosts the OpenAI models within its Azure infrastructure, and all workspace data sent as context to Azure OpenAI remains within the Azure OpenAI service.

Legal and Compliance Inquiries

Q: What does the AI terms agreement contain, and why is it necessary to agree to it?

A: AI terms simply acknowledge that LeanIX uses AI functionality to deliver certain features. It's a standard acknowledgment required to ensure transparency and it does not pose any additional obligation to the customer.

Q: Do the base AI capabilities process personal data? Should customers update their contract terms with LeanIX due to AI capability activation?

A: The base AI capabilities do not process personal data, so there is no need to update the contract's data processing terms.

Q: What agreements did LeanIX sign with Microsoft regarding these AI capabilities?

A: LeanIX has standard commercial terms with Microsoft for the use of their AI functionalities. More details can be found on Microsoft’s Azure OpenAI Service.

Q: Are there any additional terms from Microsoft Azure Open AI Services for the LeanIX base AI capabilities?

A: There are no additional terms from Microsoft. Any future updates will be made available via LeaniX Commercial.