Mohammed Ajouz, Senior Vice President, Global Head of Technical Support at SAP, joined SAP in late 2017 and has overseen a significant transformation of the company’s support model. When he arrived, customer support was built around a traditional structure, relying heavily on a searchable knowledge base and limited communication channels.
Since then, the organization has steadily evolved into a modern, technology-driven customer support operation. Ajouz highlights that AI has long been central to this evolution, beginning with machine learning (ML) and robotic process automation (RPA), and progressing through predictive, conversational, generative, and now agentic AI. Today, thousands of customers benefit from advanced tools, including Incident Solution Matching, Support Assistant, Support Log Assistant, Support Trend Analysis, Machine Translation, Categorization Predictors, and GenAI search, he said.
Ajouz recently sat down with ASUG Executive Exchange to discuss the evolution of SAP’s support model, including Agentic AI, fostering innovation across globally distributed teams and how SAP customers are benefiting. In the end, Ajouz notes: “Our drive is to reduce the customer effort and deliver an exceptional customer experience.”
How is SAP measuring the impact of the support organization? How are customers experiencing the results? Can you share an example?
Our north star is customer effort score (CES). Research shows the likelihood of customer loyalty increases with a low-effort experience. There are many other data points we monitor, one of them being the implicit deflection rate. Out of all the search sessions initiated on our portal, only 15% of them result in a case being created. Inversely, customers can find their solution using our tools and technologies 85% of the time. Combined with CES, these are great indicators of how successful we are with our promise to deliver an exceptional experience.
In what ways has AI changed how support teams prioritize and resolve issues today compared to five years ago?
AI has significantly transformed how support teams prioritize and resolve issues over the past five years. Here's a quick breakdown of the key changes:
- AI-Powered Self-Service and Deflection: Knowledge bases were static and required manual searching. Chatbots were basic, often frustrating. Now modern AI uses natural language understanding (NLU) and large language models to power intelligent virtual agents that resolve a large portion of issues without human intervention. These engines continuously learn from support interactions.
- Predictive and Proactive Support: Support was mainly reactive—issues were addressed after customers reported them. Now, AI can detect patterns and anomalies in customer systems to predict failures before they happen, and support teams can proactively resolve issues, often before customers notice them.
- AI as a Co-Pilot for Support Engineers: We have built many use cases where AI assists engineers in real time by suggesting resolutions, summarizing tickets, creating knowledge articles, and even drafting responses. This accelerates resolution time and improves accuracy.
Overall, AI has significantly transformed our capabilities to deliver faster resolution times, higher customer satisfaction, and improved support engineers’ experience.
Let's talk a bit about “Agentic AI.” How is it reshaping SAP's support model? How does this differ from GenAI from a support perspective?
In the past two years, generative AI (GenAI) has been at the forefront of technology innovation. However, just as many enterprises are still coming to grips with GenAI, we’re already shifting toward the next chapter: AI agents. Agentic AI presents another wave of transformation for customer support, offering the potential to highly automate interactions and processes.
Until now, AI models such as large language models (LLMs) have performed tasks such as generating text and summarizing documents, but they haven’t been able to act by themselves on their own “initiative.” Instead, they’ve acted on prompts. Agentic AI is changing that. It allows for the ability to select what actions to take for achieving particular outcomes. This provides a significant opportunity for performance gains that will increase over time as the systems evolve to more effectively achieve their goals. These systems can also plan and execute autonomous actions, dynamically adapting their approach based on context and ongoing analysis to achieve a specific goal.
Unlike robotic process automation, Agentic AI doesn’t require explicit inputs and doesn’t produce predetermined outputs. Agentic AI entities can receive goal instructions, iterate and delegate tasks, and output variables and dynamic information—often augmenting the user’s work.
We are not just talking about the potential of AI; we are starting to reimagine how a user’s experience with Agentic AI would be, embedding AI across our end-to-end business processes.
Please share an example where proactive support prevented a major incident (such as Black Friday).
During the 2024 Black Friday Cyber Monday (BFCM) event, SAP revolutionized support for its SAP Commerce Cloud customers through a groundbreaking preventative care model powered by AI and automation. This innovative approach marked a transformative shift from previous years' laborious, time-consuming, and inconsistent manual processes.
In past BFCM events, analyzing all customer environments and accurately assessing risks within the short preparation window was an overwhelming challenge. The process was prone to human error, inconsistencies, and often resulted in reactive firefighting during the critical sales period.
This year, leveraging AI and automation streamlined the entire process dramatically. The AI-driven system swiftly analyzed 143 customers from 57 countries across 27 industries, a task that would have taken weeks of manual effort previously. By harnessing predictive analytics and machine learning, SAP accurately assessed customer risk scores and provided targeted recommendations, ensuring proactive issue resolution.
And the outcome was that SAP Commerce Cloud recorded US$12.2B in gross merchandising value for customers, with a 23.4% YoY increase in the number of orders processed, with 100% uptime.
How do you envision the future of enterprise technical support in the next 2–3 years?
The future of support is invisible support—through prevention, monitoring, and auto-resolution. Invisible support, or predictive support as it is more commonly known, is about solving problems before the user experiences them or even knows they exist. It’s the next evolution of support, driven by data, AI, and automation. Here's why it's transformative: Invisible support uses real-time signals (system logs, telemetry, usage patterns, error rates, etc.) and historical data to anticipate issues before they occur, proactively alerting teams or users and—if allowed—automatically resolving issues or recommending actions to avoid operational impact. It shifts the entire paradigm from “I have a problem, let me contact support” to “support fixed something I didn’t even know was broken.” A few years ago, that was nearly impossible, but with the introduction of generative and Agentic AI, it is becoming more of a reality. Invisible support will become a competitive differentiator for customer experience.
Another trend we will see is embedded support. Help and guidance will be built into the product experience, surfacing just-in-time guidance without needing to contact Support or log a case. Ultimately, we will witness a shift toward a frictionless, embedded, and proactive support.
How do you foster innovation and trust across a globally distributed support team?
One of our core operating principles is to encourage innovation and creativity in pursuit of higher goals. That means creating space for the 1 percent possibility — the boldest, most imaginative ideas our team members have. To unlock those ideas, we need to foster psychological safety in the workplace. People need to feel safe sharing openly with leadership.
It starts with leadership modeling vulnerability and holding blameless retrospectives. When people know they will not be punished for failure, they are more willing to take risks and share ideas that might seem unconventional. Over time, this nurtures a habit of innovation and builds a culture of continuous improvement.
But psychological safety alone is not enough. You also need trust. Trust fuels innovation, and it grows when people feel ownership of their work and are empowered to experiment, solve problems, and collaborate across teams and regions.
Bottom line: trust is the foundation. Innovation is the outcome.
What leadership principles guide your decision-making during high-pressure situations?
Colin Powell once said, “Great leaders are almost always great simplifiers, who can cut through argument, debate, and doubt, to offer a solution everybody can understand.” I try to embody this approach in high-pressure situations by remaining focused on the purpose and cutting through the noise. People typically mirror the energy of their leader, so staying calm and collected is also key. Projecting confidence while making space provides clarity for others to think without emotional volatility. Most importantly, I trust my team.
Leadership is not about having all the answers in a crisis but about creating the conditions for the right answers to emerge.