Adopt the SAP Business AI Platform without betting your roadmap on one model or one vendor.
At Sapphire 2026 SAP pulled BTP, Business Data Cloud, and AI Foundation into one architecture: the SAP Business AI Platform. The context layer (Knowledge Graph, Domain Models, BDC), the build layer (Joule Studio), and a governance layer for tracking every agent. Mindset has been building on that stack since before it had a single name. We help you stand up the platform, ground it in your real business data, and run AI on top of it, whether that AI is SAP Joule or your own models on AWS Bedrock, OpenAI, or Anthropic. The SAP-side work is the same either way, and that work is where most enterprise AI projects quietly stall.
The reality
What changes
One grounded platform
We set up the SAP Business AI Platform so the context layer (SAP Domain Models and BDC today, with the Knowledge Graph as it matures) gives every agent a consistent map of your entities, processes, and relationships across SAP and non-SAP data.
Endorsed paths from day one
Every integration runs through Joule, Integration Suite, or the MCP Gateway, the endorsed paths SAP is steering partners toward in 2026. No rework when the guidance tightens further.
Your model or SAP's model
Run Joule with SAP Domain Models, or bring your own LLM on Bedrock, OpenAI, or Anthropic. Either way SAP grounds and serves it. The Generative AI Hub and the MCP Gateway make both routes the same SAP-side build.
Trust as an architecture choice
Agents read through the same official OData APIs Fiori uses, over Cloud Connector and BTP destinations, with no shadow data. A user can verify any answer against the system of record.
Governance you can see
Every agent registered, tracked, and monitored, with human approval required on writes and a clear audit trail for agent-initiated actions.
Critical insight
The model is not the moat. We have shipped agents on real customer landscapes and the lesson every time is the same: trust comes from the data layer underneath the AI, not the AI. Get the SAP foundation right and you can swap models without rebuilding it.
We work across the whole platform: the context layer that grounds AI, the build layer that creates it, and the governance layer that keeps it accountable. The agent-building practice has its own home on the Joule Agents page. This is the platform and data work that makes all of it possible.
01
Stand up the SAP Business AI Platform: AI Core, Generative AI Hub, and the BTP environment configured for AI workloads with identity, governance, and the right entitlements from the start.
02
Make your SAP landscape machine-readable for AI. SAP Domain Models give agents real business semantics today, so answers reflect your posting logic and master data, not a generic guess. The SAP Knowledge Graph is the emerging next layer SAP showed at Sapphire 2026, and we track it as it moves from early adopter access toward production.
03
Get AI-ready data products out of BDC, Datasphere, and SAP Analytics Cloud, with the guardrails that keep key-control reporting sourced from the SAP core. We did exactly this work at Clorox.
04
Turn S/4HANA OData APIs into tools an agent can call, scoped by business domain so the model picks the right one every time. We run this pattern in production today.
05
Already invested in Bedrock, OpenAI, or Anthropic? Keep it. We make SAP the platform that grounds and serves your models, so the SAP work pays off no matter where the model lives.
06
Find the use cases worth building through a design-thinking workshop, then put the policies, identity, and audit controls in place so AI scales without becoming a risk.
After migrating 30+ legacy integrations to BTP, Border States made Mindset its go-to partner for mobile and AI on the platform. The Gen AI Sales Assistant and quote automation work cut a multi-hour order task to minutes.
~2 hours ~2 minutes
to enter a 100+ line-item order
In a Dearborn hackathon, the Ford, SAP, and Mindset team built Currency Revaluator, an AI-assisted app on BTP that replaces a multi-day, multi-system foreign-exchange reconciliation grind with proactive anomaly detection on a single screen. The team won.
One screen
for an FX reconciliation that spanned 500+ company codes
Mindset helped Clorox build the data foundation for autonomous finance on S/4HANA, Datasphere, and SAC, with guardrails routing key-control reporting through the SAP core so the data feeding AI is clean and trusted.
Real-time
data foundation for autonomous finance
We treat AI like any enterprise integration: scope tightly, ground in real data, keep a human on writes. The fast wins are real, and they sit on a foundation built to last.
Phase 1
2 to 4 weeks
A design-thinking workshop to find the use cases worth building, score them by pain and value, and map the as-is process. We assess your SAP landscape, data readiness, and where the trust gaps are.
Phase 2
3 to 6 weeks
Configure the SAP Business AI Platform: AI Core, Generative AI Hub, BDC, identity, and the context layer. We set up endorsed paths and the grounding that makes later agents trustworthy.
Phase 3
4 to 8 weeks
Build a grounded pilot against real data, Joule or your own model, with human-in-the-loop on every write. The bar is the Friday-to-Monday test: would a business user act on this without checking?
Phase 4
Ongoing
Extend across domains with reusable patterns, register and monitor every agent, and keep the platform current as SAP's policies and models evolve.
Preeti Kumari
Sr. Delivery Analyst
Joule & Agents
I recently got an opportunity to attend SAP SIT for the first time at the SAP Campus in Whitefield, and…
Dr. Niz Safrudin
Product Innovation & Experience-Driven Transformation
Joule & Agents
In late April, SAP reopened the SAP AppHaus in Walldorf, Germany in Building 5. It’s a new space with an…
Jonathan Bragg
VP of Product - BTP, BDC, AI
Joule & Agents
Gavin's post called the "pipes" the second layer of the Mindset Stack. In this blog, I am going to detail…
Get your SAP core standard, governed, and upgrade-safe, so the AI era can actually be trusted to run on it.
Read more
An agent factory for SAP
Read more
Retire PI/PO before 2027 and build the integration layer your agents will run on.
Read more
SAP support from the people who understand the architecture, not a generic helpdesk.
Read more
The path from BW to intelligent enterprise
Read more
Ship SAP changes on demand, without putting the system of record at risk.
Read more
Presented publicly
Talks our team gave on this topic at SAP events. Slides included.
SAP put €100M (~$114M USD) behind partner-led Business AI adoption, up to €100,000 per customer. We're an approved partner and we guide you through SAP's process. The window closes at the end of 2026.
The first call is a working session, not a pitch. Bring a real problem and we’ll whiteboard it with you.