What we do
SAP AI Integration Services
Fixed-scope engagements for enterprise teams connecting SAP systems to AI agents. Every project is delivered by an engineer with 14+ years of SAP experience — not subcontracted.
MCP Connector Build
Most popularWe expose your SAP system — S/4HANA, ECC, or BTP — as a set of MCP tools that any AI agent can call. The result: Claude, GPT-4, Cursor, or any MCP-compatible agent can query your SAP data and trigger SAP processes in natural language, without custom integration code for each use case.
What's included
- ✓Discovery session to map the SAP data sources and processes in scope
- ✓MCP server built in TypeScript or Python, deployed to your infrastructure or BTP
- ✓Authentication wired to your existing SAP OAuth or basic auth setup
- ✓Tools for read and write operations (OData, RFC, BAPI — as appropriate)
- ✓Test suite and documentation for each exposed tool
- ✓Handover session with your technical team
Good for
- →Teams wanting to connect an existing AI assistant to SAP data
- →Organisations evaluating MCP before a larger platform commitment
- →Developers who want a production-grade connector without building it themselves
AI Process Pilot
End-to-endWe take one SAP process — procurement, finance, HR, logistics — and fully automate it with an AI agent end-to-end. You define the process; we design the agent, build the MCP connector, wire up the LLM, and deliver a working system your team can demonstrate to stakeholders.
What's included
- ✓Process scoping workshop to identify the highest-value automation target
- ✓MCP connector build for the relevant SAP data and transactions
- ✓AI agent development with the model of your choice (Claude, GPT-4o, or other)
- ✓User interface or integration point (chat UI, Slack bot, Teams app, or API)
- ✓Testing with realistic SAP data in a non-production system
- ✓Deployment guide and documentation
Good for
- →Teams that need a working demo to secure internal buy-in
- →Organisations with a specific high-value process they want to automate
- →IT leaders who want to validate the approach before committing to a full rollout
Architecture Advisory
StrategicStructured guidance for SAP AI architecture decisions. We review your current landscape, assess your readiness for the SAP Business AI Platform and MCP, and produce a practical roadmap that avoids the integration patterns that will need to be reworked in 18 months.
What's included
- ✓SAP landscape review (systems, versions, integration patterns in use)
- ✓Assessment against SAP Business AI Platform architecture (MCP, A2A, Agent Gateway, AI Core)
- ✓Identification of the two or three highest-value AI automation opportunities
- ✓Roadmap with sequenced delivery phases and realistic timelines
- ✓Technology selection guidance — Joule vs custom agents, model choices, hosting options
- ✓Written deliverable summarising findings and recommendations
Good for
- →CIOs and IT directors planning SAP AI strategy for 2026–2027
- →SAP programme managers handed an AI mandate without a clear starting point
- →Teams that have started a project and want an independent technical review
Ready-made tool
Don't need a full engagement?
SAP ADT MCP
Connect Claude, Cursor, or any MCP-compatible AI agent directly to your SAP ABAP Development Tools. Browse packages, read objects, search code — from your AI assistant. Works with S/4HANA and ECC. TypeScript source included.
Get it on GumroadCommon questions
FAQ
Do we need to be on S/4HANA or can this work on ECC?
Both work. MCP connectors call SAP via RFC, BAPI, or OData — all of which are available on ECC and S/4HANA. Being on BTP opens additional options (AI Core, Joule Studio), but it is not a prerequisite.
Does our SAP data leave our network?
Only if you choose to send it to an external LLM. We can architect the solution to keep all data within your network using locally hosted models, or we can use SAP AI Core which runs models within SAP's governed BTP infrastructure.
How is this different from what SAP Joule does?
Joule is SAP's built-in AI assistant, best suited to standard SAP processes within the SAP ecosystem. MCP-based custom agents are model-agnostic, cross-system, and fully customisable. We help you choose the right approach — and often the answer is both.
What is the minimum engagement?
The MCP Connector Build starts at €5,000 and delivers a production-grade SAP MCP connector in 2–3 weeks. This is designed as an entry point — a low-risk way to validate the approach before a broader rollout.