SAP LeanIX now allows you to access your enterprise architecture information using AI prompts in Claude or Microsoft Copilot. Discover truly AI-native enterprise architecture with SAP LeanIX.
We're excited to announce the release of MCP apps within our MCP server. This update allows you to see your SAP LeanIX fact sheets, relationship maps, reports, and dashboards directly within your Claude or Microsoft Copilot chats.
Starting within a chat interface, your prompts can open in-line interfaces to let you work in your instance SAP LeanIX without having to leave your AI chat interface. Open fact sheets and drill into dependencies securely while you work within an AI environment.
This new capability turns SAP LeanIX into an AI-native enterprise architecture solution. Leverage our enterprise architecture AI agents to retrieve intelligent inventory data for you based on the context you provide.
Book a demo of MCP apps today, or read on to learn more about why we've built this powerful capability:
The three challenges of custom AI integrations
When companies first started connecting AI agents to business tools in 2024, most took the fastest path: building custom integrations through bespoke API wrappers or homegrown model context protocol (MCP) connectors.
This shortcut created three critical problems:
1 No schema
With a custom integration, you don't have a standardized schema and AI agents treat your architecture data as generic API responses. They misinterpret relationships between fact sheets, confuse application types with IT components, and return answers that look plausible. but violate your business logic.
2 No governance
Direct API integrations or homegrown MCPs may bypass governance controls. Agents may access and over-ride fact sheets outside a user's authorized scope or leave no audit trail of what data was accessed or whether it was exposed to unauthorized users.
3 No standardization
Lastly, every new AI platform (Claude, Copilot, Gemini, custom LLMs) requires its own integration code. Every change then demands re-engineering all your custom connectors to ensure agents can access and interpret the new capabilities.
MCP server: democratizing access to EA data
To allow non-technical users to access EA data with natural-language prompts without facing the three challenges above, we launched the SAP LeanIX MCP server in 2025. This provides agents with a schema for reliable semantic understanding of enterprise architecture data, so agents identify the right inventory components and interpret attributes, relations, decisions and road maps correctly.
MCP server security is enforced through OAuth authentication. Technical users can, therefore, fully explore architectural domains, reports and decisions while business users only have access to data relevant to their role.
The MCP server ensures cross-platform compatibility with major agentic platforms. It ensures features and meta model changes are accessible in agentic conversations, meaning you don't need to worry about technical maintenance or overhead.
However, this wasn't a full AI-native EA solution until now.
Breaking the text wall in agentic conversations
When users connect AI agents to EA data, something can be lost in translation. Insights that were immediately clear in visual form just become walls of text and details can be lost, such as business criticality score and custom attributes.
The SAP LeanIX MCP apps extension for the MCP server, however, turns text responses into visual components that anyone can understand and interact with. Without leaving the AI interface, you can benefit from the visual clarity of SAP LeanIX.
For example, if you ask "what depends on our AC management system?", you'll see a clickable relation explorer map with access control (AC) management at the center, instead of just a text list of application names. Each connected node represents a dependency to custom-built applications, interfaces, and data objects.
MCP apps go beyond first-level connections, however. In conversation, you can click on any dependent application, and the map expands to reveal second and third-level dependencies:
-
Active projects currently modifying that application
-
Geographic regions where the application is deployed
-
Business capabilities that would be disrupted if it’s replaced
-
Organizational teams responsible for maintaining it
This multi-dimensional view doesn't just answer the question of "what's connected?" It shows you the full business impact of changing your AC Management.
If an application looks unfamiliar, you can explore it further as a visual fact sheet without leaving the chat. Attributes like business and functional fit, and lifecycle data provide the context for making decisions on the spot and informing the right teams.
MCP apps in action: supply chain security response
The true power of MCP apps is in accelerating your response time in critical use cases, like efficiently planning mitigations to cybersecurity threats using live inventory data and visual embeddings. For example, if an enterprise architect needed to assess the impact radius of a supply chain attack compromising a widely-used LiteLLM version, they would need to know if this a low-priority cleanup or an executive-escalation incident requiring action within the hour.
Switching between different applications and AI interfaces takes time. Now, however, you can navigate through different software bills of materials (SBOMs) and piece together which AI agents use the compromised version, all in one interface.
MCP apps generate custom visualizations on demand, tailored to your specific question. The agent then analyzes your SBOMs, identifies affected applications, and renders an interactive impact map showing which agentic applications depend on the vulnerable LiteLLM version, all within the conversation.
AI-native enterprise architecture goes beyond UI
The addition of MCP apps to the SAP LeanIX MCP server is not a product upgrade; it's a shift to AI-native enterprise architecture. Through agents and the MCP server, it facilitates a direct connection, expressed in natural language, between business intent and intelligent data, creating an unrestricted mode of interaction optimized for exploration, analysis, and decision-making.
Moving forward, the enterprise architecture inventory you've built in SAP LeanIX remains your source of truth. MCP server with embedded MCP apps, however, makes it accessible and visually explorable in AI conversations.
This is AI-native enterprise architecture and it's now available in SAP LeanIX. To find out more, book a demo today:
