Based on “Beyond Davos Hype: Real-World AI Challenges and their impact on SAP” – A LinkedIn Live Panel by LeapGreat.
If AI Does the Junior Work, Who Leads Your SAP Transformation in 10 Years?
AI can generate an S/4HANA business case in seconds.
It cannot lead a steering committee. It cannot align finance and operations. It cannot own a failed cutover.
At the 2026 World Economic Forum in Davos, leaders warned about “cognitive atrophy.” That risk now applies to SAP transformation programs.
📺 Watch the Full Panel Discussion
Title: Beyond Davos Hype: Real-World AI Challenges and their impact on SAP
Date: February 26, 2026
Duration: ~60 minutes
Panelists:
- Stefan Schoepfel, CEO, Value AI Institute
- Bernhard Lang, Co-founder, LeapGreat
What the Discussion Covered
During the session, Stefan Schoepfel and Bernhard Lang examined what has changed in SAP delivery and what has not.
They demonstrated how agentic AI connected to structured process architecture can:
- Generate full project dashboards in minutes
- Complete process validation in about one minute
- Produce executable test cases automatically
They also addressed the leadership implications of automation, the decline in junior hiring, and the emergence of the “AI Fixer” role.
Key Takeaways: Beyond Davos Hype
1. Enterprise AI Is Still Early
Personal productivity AI is widespread. Business AI embedded inside core processes is not.
Most enterprises are running pilots. Few have scaled AI across functions. Even fewer have integrated it into SAP-driven operations.
Large global surveys confirm this. Enterprise-level financial impact remains limited. Agentic AI adoption is rare .
There is still room to build structured capability. There is not room to ignore the shift.
2. SAP Delivery Has Stagnated. AI Changes the Math.
The mechanics of SAP implementation have barely moved in 20 years. Documentation is manual. Process validation is manual. Test case creation is manual.
AI compresses this layer.
LeapGreat demonstrated:
- Full project dashboards generated in minutes
- Process validation reduced from hours to about one minute
- End-to-end test cases produced in one to two minutes
With mainstream support for SAP ECC ending in 2027 and migrations running 12 to 24 months, timing matters .
AI does not eliminate transformation work. It eliminates low-value friction.
3. AI Agents Require Structured Foundations
Generative AI trained on public data is unreliable inside enterprise systems. Without structured process architecture and clean data, agents hallucinate.
Analyst forecasts warn that a large share of AI projects will be canceled due to weak data and unclear value .
The constraint is simple:
Structure first. Agents second.
If your SAP landscape is inconsistent, AI will amplify the inconsistency.
4. Productivity Gains Shift the Bottleneck
The session cited measurable gains:
- 30%+ improvement in implementation work
- 50%+ compression in strategy and benchmarking tasks
Independent studies report similar ranges .
But AI delivers an 80% solution. Humans validate, adjust, and decide.
Execution speeds up. Leadership becomes the constraint.
The real value is time reallocation. When documentation and test scripts are automated, architects and business leads focus on:
- Process redesign
- Target operating models
- Governance
- Commercial differentiation
Organizations that reinvest freed capacity into higher-value work outperform those that treat AI as a cost reduction.
5. The Post-Execution Economy Requires New Roles
If AI handles structured execution, human value shifts upward.
The session defined a new capability stack:
- Vision
- Architecture
- Orchestration
- Change leadership
- Governance
- Judgment
This is where “cognitive atrophy” becomes a risk.
Seniority develops through repetition. Exposure to small problems builds pattern recognition. If AI absorbs those problems and junior hiring collapses, the next generation of SAP leaders never forms.
Data shows entry-level hiring is declining sharply, even as younger cohorts demonstrate higher AI readiness .
Cutting junior roles may improve short-term cost metrics. It weakens long-term capability.
Young hires are not overhead. They are leadership infrastructure.
6. The “AI Fixer” Is Now Critical
Most AI initiatives stall between pilot and production.
The missing role is not a data scientist. It is a translator.
The AI Fixer bridges:
- Business process owners
- SAP architecture
- Data governance
- Security
- Change management
This person moves AI from experiment to production.
Without this bridge, AI stays in PowerPoint. With it, systems ship.
Every serious SAP program will need this capability.
7. Consulting and Delivery Models Are Shifting
AI compresses research cycles and reduces manual configuration work. Large firms are investing heavily. Boutique firms are combining AI with high-judgment delivery to challenge traditional models .
The premium shifts from labor scale to decision quality.
Clients will pay for outcomes. They will not pay for slide volume.
SAP + Microsoft: Real-World Constraints
Integration between SAP S/4HANA, Joule, and Microsoft Copilot is now generally available.
The constraint is not branding. It is governance.
Identity design, role architecture, and data exposure rules determine whether Copilot enhances productivity or creates risk.
Integration is technical. Adoption is organizational.
The Leadership Risk
AI will not eliminate SAP transformation roles.
It will eliminate leaders who cannot use AI and cannot build teams that use it.
If AI performs the junior execution layer, organizations must deliberately design development pathways for future architects and program leads.
Ten years from now, someone must:
- Set the direction
- Validate the architecture
- Manage the cutover
- Own the outcome
AI can assist each step. It cannot carry accountability.
Davos warned about cognitive atrophy. In SAP transformation, that risk is not theoretical.
If you remove the work that builds judgment, you remove the leaders who own transformation.
The organizations that treat AI as a capacity multiplier and a leadership accelerator will build durable advantage.
The ones that treat it as labor substitution will discover the gap at their next go-live.
LinkedIn Live Transcript: Beyond Davos Hype: Real-World AI Challenges and their impact on SAP
Below is the full transcript from the LinkedIn Live session, lightly edited for clarity and formatting.
Bernhard Lang: Hello, my name is Bernhard Lang. I’m your host today and Co-Founder of LeapGreat. I’m very happy to see so many people joining.
At the beginning, a few housekeeping notes. We have a chat, so please use it. Ask your questions, share comments, what you like, what you dislike. We will answer questions at the end of the session. We want this to be interactive.
Also, let us know where you’re joining from. Drop it in the comments so we can see who is dialing in from the furthest away.
I would like to welcome my guest today, Stefan Schöpfel, CEO of the Value AI Institute. Stefan is a long-time SAP veteran. He was formerly Global Vice President responsible for Business AI go-to-market and spent many years at SAP in the Intelligent Enterprise Institute and analytics space. A perfect person to discuss today’s topic.
We titled this session Beyond the Davos Hype because Stefan and I met in Davos a few weeks ago during the World Economic Forum. Stefan hosted a full-day session focused on AI and Business AI.
Today we want to talk about real-world AI challenges and their impact on SAP transformation.
Stefan, welcome. Great to have you here.
Click to view the full transcript.
Stefan Schoepfel: Thank you, Bernhard. Very happy to join.
Let’s start with the question: What is Business AI, and how does it compare to generative AI?
There is a lot of hype around generative AI tools. Most of them are trained on web data, such as Wikipedia, Reddit, and other public sources. They are not deeply trained on enterprise data.
When we speak about Business AI and moving from hype to real outcomes, the key question becomes: How can models be trained on company data? How can they be embedded in real business processes?
If we are honest, the AI journey in enterprises is still at the beginning. Personal productivity use cases are advancing quickly. But AI embedded into end-to-end logistics chains, finance processes, or transformation programs is just starting.
That requires good data, structured processes, and trust between enterprises and technology providers.
Bernhard Lang: That is very interesting. Let’s talk about transformation.
Company transformations require heavy business involvement. Users are already busy. There is a lot of routine work: documentation, testing, validation.
If AI can offload routine work in transformation programs, that would be powerful.
There are two dimensions here:
First, AI accelerating implementation work.
Second, AI helping re-engineer business processes, not just lift-and-shift, but redesign.
How do you see AI’s role in SAP transformation projects specifically?
Stefan Schoepfel: I see huge potential.
SAP delivery has not fundamentally changed in 20 years. Documentation, training materials, test case writing—much of that is still manual.
AI can accelerate all of this.
It can support creative design decisions, but it can also eliminate routine tasks like documentation, training video creation, and test case writing.
For customers, this is a multidimensional opportunity. Migration to S/4HANA is a major effort and investment. If AI reduces the burden, more customers will move faster.
Once they are on cloud platforms, they can benefit from embedded AI capabilities.
Bernhard Lang: Let me share a few concrete examples from LeapGreat.
Our vision has always been to accelerate SAP implementations and create structured documentation. We built what we call the LeapGreat Hub, a structured digital twin of business processes and application architecture.
This structured foundation works very well with agentic AI.
For example, in a project with 521 processes and 35 RICEFs, we generated a complete dashboard without coding in about two minutes. It provided full transparency into process status, testing, refinement, and defects.
In another example, we ran 14 validation rules against a “Sell from Stock” process. Normally, this takes about two hours manually. The AI completed it in about one minute and flagged discrepancies automatically.
We also generated executable end-to-end test cases from validated processes within minutes. Traditionally, this takes several hours and must be updated manually when processes change.
We estimate more than 30% efficiency improvements in implementation projects compared to traditional approaches.
We are also able to auto-generate RFPs for full transformation programs.
Stefan Schoepfel: This is a strong use case for Business AI.
Few people enjoy writing documentation or RFPs. If AI can handle that work, it frees time for higher-value activities like redesigning processes and differentiating business models.
That is where value is created.
Bernhard Lang: You also mentioned the “post-execution economy.” Can you explain what that means?
Stefan Schoepfel: Historically, value in organizations came from execution skills. As AI becomes strong at execution, human value shifts upward.
Future skills include:
- Vision
- Architecture
- Orchestration
- Change management
- Governance
- Judgment
AI can execute. It cannot set direction. It cannot take accountability.
We may be the last generation that manages only human teams. The next generation will orchestrate hybrid teams composed of humans and AI agents.
This requires new capabilities.
Bernhard Lang: There is also concern about junior roles. Many companies are hiring fewer entry-level resources. How do you see that?
Stefan Schoepfel: AI readiness is low in most organizations. Hiring digitally native talent with AI skills can actually increase readiness.
If students learn how to orchestrate AI agents, design human-agent workflows, and think end-to-end, they are not simply junior resources. They are value contributors.
The key is integrating AI learning into education and continuous professional development.
Bernhard Lang: The pace of change feels very fast. How do you see it evolving?
Stefan Schoepfel: The pace is accelerating. There are advancements weekly and new capabilities monthly.
That makes lifelong learning essential.
Everyone working in this space should integrate learning into their weekly routine: testing tools, benchmarking use cases, experimenting with workflows.
We have seen technology shifts before, such as client-server, cloud, and e-commerce. This shift is larger.
It requires governance and caution. But it also brings significant opportunity.
Bernhard Lang: That is a good closing thought.
Stefan, thank you for the conversation. I look forward to working together on real customer projects.
Thanks to everyone who joined today. Stay tuned for upcoming LeapGreat Live sessions.
Have a great evening and thank you for joining.
