Ask why an invoice failed to post, trace an IDoc through the chain, follow an integration message end-to-end, answer a business question from live SAP data, or modernise a legacy Z object, all through one audited platform. People receive clear, human-readable answers covering exactly what their own SAP user is authorised for, and nothing beyond it. Governance is enforced in code, not in prompts.
Diagnose live issues in minutes: short dumps with source extracts and call stacks, gateway errors, failed IDocs, stuck batch jobs, system messages and integration message logs, all explained and correlated in plain language. No SAPGUI archaeology, no waiting on a developer.
Safe, self-service answers from live SAP data: did an invoice post, where is a purchase order, what are this year’s totals by account. Aggregations run inside SAP, and a role-based policy engine keeps these users strictly read-only, so self-service never becomes uncontrolled access.
Understand, validate, change and modernise the custom estate: from explaining undocumented Z code and tracing dependencies, to governed ABAP changes, to scaffolding a classic Z-report into a modern RAP application with a live OData service and a generated Fiori app.
One audited tool surface for diagnostics, live data, integration and engineering, available from the platform's own chat UI and, via the open Model Context Protocol, from AI tools your team already uses. Identical governance and identical audit entries on every path.
Every call is scoped to the individual user’s own SAP authorisations, and the controls are enforced in code, not in prompts. The AI never decides what a user may see or do.
Every proposed change is validated with syntax and ABAP Test Cockpit checks before anything is written. Releasing a transport towards test and production, or deleting an object, remains an explicit human action the AI cannot invoke.
Every interaction, read or write, lands in a complete, append-only audit record: the real user’s identity, the parameters, the SAP response, and the decision of every governance gate.
Customer-hosted on SAP BTP, Kubernetes, Docker or air-gapped. SAP data never transits ETZ infrastructure. You bring your own AI provider, so model choice and data residency stay in your hands.
FAQ Questions
If your question isn't covered here, reach out and our team will get back to you.
Start a ConversationA governed AI runtime for your SAP landscape. People talk to their SAP system in everyday language and receive clear, human-readable answers, covering exactly the information and capabilities their own SAP user is authorised for, and nothing beyond it. It serves the whole SAP team: operations and support, functional consultants, business users, and developers.
The two are complementary. Joule assists business users inside SAP applications; the ETZ AI Platform is a governed AI runtime for the teams that run, support, build and modernise the SAP landscape itself.
The platform is customer-hosted. Deploy it on SAP BTP Cloud Foundry, on Kubernetes or Docker inside your own network, or fully air-gapped. SAP data never transits ETZ infrastructure. The AI provider is your own account (Anthropic, OpenAI, or Azure OpenAI), so model choice and data residency remain yours.
Every action the platform performs in SAP runs under a real, named SAP user, never a shared technical super-user. On SAP BTP the platform supports per-user principal propagation, so SAP’s own logs show the actual person. A scope-aware policy engine constrains what each role can do on top of that.
No. Writes are structurally impossible outside a development system, and every change must pass syntax and ABAP Test Cockpit validation before anything is written. Test and production stay reachable only through transport release: the human sign-off moment SAP teams and their auditors already know, and one the AI cannot invoke at all.
Diagnose incidents across the SAP stack, answer questions from live SAP data with in-system aggregation, operate SAP Cloud Integration end-to-end, explain and validate custom ABAP, execute governed changes, and scaffold classic Z objects into the full RAP stack with a live OData service and a generated Fiori Elements app.
Live-verified against SAP S/4HANA on-premise and private cloud, SAP BTP Cloud Foundry, and SAP Integration Suite (Cloud Integration). The platform talks to SAP through standard ADT, OData and RFC interfaces. A published compatibility matrix covering earlier releases is in progress; ask us about your specific landscape.
Yes, and the platform is built on that assumption. AI-generated answers and code proposals must be reviewed by a qualified human before being relied on. That is why every change is validated against SAP’s own checks before it can be written, why writes only ever reach development systems, and why promotion to test and production remains an explicit human decision.
We have a strong global presence, marking our commitment to delivering tech excellence in key markets worldwide.