Cloud SAP® automation now has a standards-based path to AI

Key Takeaways

• Connecting AI assistants to SAP requires custom integration work that must be rebuilt for every new AI platform the organization adopts.
• The Precisely Automate MCP Server exposes existing Automate Studio scripts and Automate Evolve workflows to any MCP-compatible AI tool through one standards-based connection.
• Native OData support in Automate Studio ensures cloud SAP automation runs through SAP-sanctioned APIs, not legacy scripting approaches now restricted under SAP connectivity policies.

I’ve noticed a trend among SAP customers: they’re invested in AI tooling, their teams are using Claude, Copilot, or ChatGPT daily, and they’re making very real productivity gains in content, reporting, and analysis. Then someone asks the natural next question: can my chosen AI do something in SAP?

The answer is almost always, “not yet,” or “not very well.” Compounding the challenges, SAP’s updated API policy requires customers to direct AI access to SAP-published APIs through endorsed pathways. Consequently, connecting an AI assistant to your SAP ERP sytem can be tricky, and given the recent policy changes, often requires custom integration work that needs to be repeated from scratch for every new AI platform your organization adopts. As more AI tools enter the enterprise stack, that debt accumulates faster than most IT teams can address it. 

Now, you can tackle that problem head-on with the new Precisely Automate MCP Server and native OData support in Precisely Automate: two capabilities that together give any MCP-compatible AI assistant governed, standards-based access to SAP automation workflows, without rebuilding the integration every time the AI stack changes.

What are the challenges of manual AI-to-SAP connectivity?

When a finance lead wants their AI assistant to trigger a journal entry workflow, or a supply chain manager needs it to update a vendor record, someone has to build the bridge between the AI tool and SAP. That means custom API development, authentication configuration, and ongoing maintenance. Then, when the organization adds the next AI platform, they build it again.

Most enterprises aren’t converging on a single AI tool, but instead running several in parallel: Copilot for productivity, Claude for analysis, ChatGPT for specific workflows. Joule for SAP tasks.

Each represents its own integration project if it needs to connect to SAP. That’s an untenable backlog for IT teams already stretched by S/4HANA migration work.

SAP’s evolving connectivity policies for third-party software and LLMs have made the stakes higher still. Organizations relying on legacy scripting approaches (BAPI, RFC, GUI automation) are finding that those methods face increasing restrictions in cloud environments, particularly S/4HANA Cloud and RISE with SAP.

SAP’s direction is clear: integrations need to run through sanctioned, API-based protocols. What was once a best practice is quickly becoming a compliance requirement.

Research from Precisely and ASUG how widespread this pressure is:

• 62% of organizations cite business process complexity as their top automation challenge
• 59% are already fully or partially live on S/4HANA.

Organizations that can connect AI to SAP through a standards-based interface will have a meaningful advantage over those still managing it integration by integration.

What does the Precisely Automate MCP server do, and how does it work?

The Precisely Automate MCP Server is built on the Model Context Protocol (MCP), an open standard developed by Anthropic and now adopted broadly across the AI ecosystem. It exposes Automate Studio scripts and Automate Evolve workflows as AI-callable tools, so almost any MCP-compatible AI client can discover and invoke them through a single configuration.

• A finance operations lead can instruct their AI assistant to run an existing journal entry automation.
• A supply chain manager can ask it to retrieve or update vendor records.
• A utilities operations lead can request updates to equipment records, maintenance orders, and plant data

All of this is achieved in plain language, without filing IT tickets or waiting on developer resources. Automate Studio and Evolve handle the SAP interactions in the background, with governance controls intact throughout.

The Studio scripts and Evolve workflows that the MCP Server exposes need to be built and configured  first. This isn’t a zero-configuration shortcut to cloud SAP automation, but what it eliminates is the recurring overhead of re-exposing those workflows to each new AI platform. Build your automation once; connect it to any MCP-compatible AI tool going forward.

Claude, ChatGPT, Microsoft Copilot, Joule and Gemini are all MCP-compatible today. As more AI platforms adopt the standard, they connect to the same Automate MCP Server without additional integration work. That compounds the value of every automation already built in Studio or Evolve, rather than treating each new AI tool as a new integration project.

This release is part of a broader strategy: Precisely is rolling out enterprise MCP Server support for multiple products across the portfolio, connecting AI agents to data quality, location intelligence, master data management, and automation capabilities through a unified interface.

Why does cloud SAP automation need native OData support?

SAP has been moving its connectivity policies in a clear direction for some time, and recent changes have accelerated that. Direct BAPI calls, RFC-based scripting, and GUI automation are increasingly restricted or unsupported for third-party software and LLMs in SAP’s cloud environments. If your cloud SAP automation strategy is built on those approaches, you’re presented with a very real risk.

OData (Open Data Protocol) is SAP’s preferred REST-based API standard for modern integrations. It’s the approach required for Clean Core environments, including RISE with SAP and S/4HANA Cloud, and it’s what SAP endorses for all new third-party connectivity. Native OData support is now built into Automate Studio, which means AI-driven SAP automation built on Automate runs through SAP-sanctioned protocols from the start.

The scope of what’s reachable through OData is substantial. The SAP Business Accelerator Hub publishes more than 2,700 OData APIs across SAP’s cloud application portfolio, covering S/4HANA Finance, SAP SuccessFactors, SAP Ariba, SAP Concur, SAP Customer Experience, and more. That breadth means your Automate-driven cloud SAP automation can extend across the full SAP application landscape, not just the on-premises processes where legacy scripting worked.

In practical terms, native OData support delivers:

• Stable automation across SAP upgrades, with no dependency on screen scraping or brittle GUI scripting that breaks with interface changes
• A full audit trail on every AI-initiated SAP transaction, which is a hard requirement for organizations under FDA, GxP, SOX, or similar regulatory obligations
• No-code data creation, retrieval, and update for business users, consistent with the governance model Automate has always delivered, now extended to OData-enabled SAP applications
• Backward compatibility with existing Studio scripts and RFC-based automations, so organizations can adopt OData for net-new automation without replacing what they’ve already built

AI-accessible and Clean Core compliant used to be competing requirements. This release makes them achievable through the same platform.

What does this mean for SAP teams building toward AI-ready data?

For autonomous systems to be useful in an SAP environment, they need reliable, governed access to the processes and data SAP holds: financial records, supply chain operations, HR, procurement. Without that access, the AI operates around SAP rather than within it.

AI systems must be able to both read from and write back to authoritative systems of record, and the Automate MCP Server extends that reach into SAP: AI agents can invoke SAP workflows through governed Automate processes, with every action logged and traceable.

If you’re already running Automate Studio or Evolve, this release converts the automation you’ve already invested in into an AI-native asset. The scripts, workflows, and governance configurations you have in place become reachable through the AI tools your teams use every day, without re-platforming.

If you’re approaching cloud SAP automation for the first time through an AI-first lens, the Automate MCP Server offers a starting point that doesn’t require committing to a single AI provider.

Our research with ASUG found that 75% of SAP organizations rate no-code and low-code capability as important when choosing an automation platform. Controlling SAP processes through natural language is the next step in that direction, and the connectivity layer to support it is now available.

Get started with a customized demo of the Precisely Automate MCP Server today.

 

Frequently Asked Questions

What is the Precisely Automate MCP Server?

The Automate MCP Server uses the open Model Context Protocol (MCP) to expose Automate Studio scripts and Evolve workflows as AI-callable tools. Any MCP-compatible AI assistant, including Claude, ChatGPT, and Microsoft Copilot, can invoke SAP automation capabilities through natural language without custom per-platform integration development.

Does the Automate MCP Server work with AI platforms other than Claude?

Yes. Because it is built on the open MCP standard, it works with any MCP-compatible client. Claude, ChatGPT, Microsoft Copilot, and Gemini are all compatible today, and any future platform that adopts the standard connects to the same Automate MCP Server without additional configuration.

What is OData, and why does it matter for cloud SAP automation?

OData (Open Data Protocol) is SAP’s preferred REST-based API standard for third-party and AI connectivity in its cloud application suite. Native OData support in Automate Studio means new SAP automation runs through SAP-sanctioned protocols rather than legacy scripting approaches that face increasing restrictions under SAP’s evolving connectivity policies for LLMs and third-party software.

Do I need to rebuild my existing Automate Studio scripts to use the MCP Server?

No. Existing Automate Studio scripts and Evolve workflows, including RFC-based automations, continue to work. The MCP Server exposes them to AI tools through a new interface without requiring a rebuild. Organizations can adopt native OData support for net-new automation where Clean Core compliance is required, while leaving existing automation in place.

What governance controls apply when AI triggers SAP automation through the MCP Server?

Governance controls remain fully intact. AI-initiated SAP actions run through Automate Studio and Evolve with the same audit trail, role-based permissions, and traceability that governs all Automate-driven processes. For organizations in regulated industries, that auditability is a baseline requirement, and it’s maintained regardless of whether a human or an AI agent initiates the workflow.

SAP automation has always been one of the higher-stakes places to get AI right. The workflows involved touch financial records, supply chain integrity, HR data, and regulatory compliance. The Automate MCP Server and native OData support bring those workflows into reach for the AI tools organizations are already running, through a connectivity approach that holds up as both the AI ecosystem and SAP’s own architecture continue to evolve.