SAP Connect Day Southern Africa brought nearly a thousand delegates to the Kyalami Grand Prix Circuit on 21 July. As sponsors, we spent the day in conversation with clients, partners and peers about what the Autonomous Enterprise means for companies in South Africa.
This was the regional follow-through from Sapphire in Madrid, so the strategic message was familiar. What made the day worthwhile was the local proof. Less roadmap, more evidence, with South African organisations describing their journeys and what they have already built.
Nazia Pillay, MD for southern Africa at SAP, opened by taking on the argument that agentic AI spells the end of the ERP era. Her framing was the opposite. "ERP is no longer just a system of record," she told delegates. "With agentic AI, we can ensure it runs end-to-end business processes, creating autonomous ERP."
The supporting numbers set the pace of change. Citing Oxford Economics research, she noted that AI now plays a part in around 30% of tasks in the average business, a share projected to reach 48% within two years.
Marc Nolla, vice president for customer advisory in EMEA at SAP, gave the statistic that landed hardest with us. Around 40,000 SAP customers have already adopted SAP Business AI in their productive systems. That reframes the conversation. This is no longer a question of whether the technology is ready.
Nolla set out the Autonomous Enterprise vision in three parts. Joule as the single engagement layer that brings teams along, SAP Autonomous Suite embedding agents and best practice across the portfolio and SAP Business AI Platform giving customers the ability to extend agents while maintaining context, reasoning and governance.
He went further to add that SAP's new agent-led transformation approach could reduce the effort required for multi-year ERP migrations by 35%. This would be a game changer for those embarking on their S4 migrations in the next 12 months.
On governance, he was direct. With Joule Studio letting workers create agents on the fly, hundreds or even thousands of agents inside a single business becomes a real prospect and the AI Agent Hub is SAP's answer: one place for IT to discover, manage and govern them all. He also flagged that enriched agentic tooling for data management, configuration, development and testing is due within the next few months, aimed squarely at improving migration projects.
The session with the most direct relevance to our own clients came from Vanessa Padiachee, Executive: Group CIO for Corporate Functions at Standard Bank Group.
She described a finance organisation at a crossroads in 2022. "We had to remediate for an end of life, end of support problem and we also had a unique business problem we needed to solve for." Since then the group has migrated group reporting and the ledgers and has completed two successful month-ends.
On lessons learned, she said one of the most significant ones was the complexity of Unicode conversion — in hindsight it should have started much earlier, it proved more complex than anticipated and became a critical dependency, and investing in foundational technical readiness would have significantly reduced downstream project risk.
Two additional points stood out. The first was her insistence on sequencing. "I am a great believer in architecture: that builds a roadmap about how you want to transform. You can unpack the roadmap and incrementally deliver against it." The second was where she now considers the industry to be. In her view, AI is no longer just around experimentation and production use cases; organisations are now going beyond that to look at enterprise scaling casesfor real business value."
On innovation she was equally clear that teams need room to experiment, but inside a controlled environment, partnering with vendors and keeping governance and guardrails in place. For anyone running a regulated institution, that balance between complexity on one side and platform on the other will be a familiar tension. She also said change management matters as much as the technology implementation itself and it's about keeping people calm during turbulence.
Stanley Dube, Head of Solution Advisory at SAP, gave the day its most quoted phrase. He described the "enterprise paradox": the expectation of instant enterprise AI colliding with the operational reality of complexity, manual processes and human behaviour.
He also gave it its most memorable staging, sharing the platform with a robot co-presenter in a tracksuit carrying Joule branding. "I have a confession to make, I have been afraid of AI," he said. "The fear I have had is that AI is going to take my job." The demonstration that followed worked through how organisations redesign jobs and processes, hold governance and build advantage.
We had the good fortune of a well-positioned stand, which served coffee in the morning, drinks in the afternoon and became the gathering point for prizes at the close. That gave us a full day of unfiltered conversation.
Three themes came up repeatedly.
Is Joule genuinely more than a chatbot? This was the single most common question. People have seen the demo. They want to know what it does on a Tuesday afternoon in a real finance function.
What about those of us still on premise? A lot of the vision assumes cloud. Delegates running on-premise estates wanted to understand which parts of this apply to them now and which parts require a migration they have not yet scheduled.
Data, again. Every conversation eventually arrived at the same place. Nobody disputed that the data foundation determines whether any of this works, which is a notable shift from a few years ago when data was the thing everyone skipped past.
If a client asked us on Monday morning what this all means for them our answer would be simple.
The direction SAP is taking is genuinely powerful and it needs to be weighed against your own context rather than adopted wholesale. Most organisations already hold more capability than they are using. Getting a clear picture of what you have entitlement to, what you are actually running and what is realistically possible from there is the most sensible first move on any AI journey.
That is unglamorous advice. It is also, on the evidence of a day spent listening to people who have done this, the advice that separates the organisations making progress from the ones still running proofs of concept.
Our thanks to SAP Africa and the African SAP User Group for a well-run day and to everyone who stopped by the stand.
SAP Connect Day Southern Africa brought nearly a thousand delegates to the Kyalami Grand Prix Circuit on 21 July. As sponsors, we spent the day in conversation with clients, partners and peers about what the Autonomous Enterprise means for companies in South Africa.
This was the regional follow-through from Sapphire in Madrid, so the strategic message was familiar. What made the day worthwhile was the local proof. Less roadmap, more evidence, with South African organisations describing their journeys and what they have already built.
Nazia Pillay, MD for southern Africa at SAP, opened by taking on the argument that agentic AI spells the end of the ERP era. Her framing was the opposite. "ERP is no longer just a system of record," she told delegates. "With agentic AI, we can ensure it runs end-to-end business processes, creating autonomous ERP."
The supporting numbers set the pace of change. Citing Oxford Economics research, she noted that AI now plays a part in around 30% of tasks in the average business, a share projected to reach 48% within two years.
Marc Nolla, vice president for customer advisory in EMEA at SAP, gave the statistic that landed hardest with us. Around 40,000 SAP customers have already adopted SAP Business AI in their productive systems. That reframes the conversation. This is no longer a question of whether the technology is ready.
Nolla set out the Autonomous Enterprise vision in three parts. Joule as the single engagement layer that brings teams along, SAP Autonomous Suite embedding agents and best practice across the portfolio and SAP Business AI Platform giving customers the ability to extend agents while maintaining context, reasoning and governance.
He went further to add that SAP's new agent-led transformation approach could reduce the effort required for multi-year ERP migrations by 35%. This would be a game changer for those embarking on their S4 migrations in the next 12 months.
On governance, he was direct. With Joule Studio letting workers create agents on the fly, hundreds or even thousands of agents inside a single business becomes a real prospect and the AI Agent Hub is SAP's answer: one place for IT to discover, manage and govern them all. He also flagged that enriched agentic tooling for data management, configuration, development and testing is due within the next few months, aimed squarely at improving migration projects.
The session with the most direct relevance to our own clients came from Vanessa Padiachee, Executive: Group CIO for Corporate Functions at Standard Bank Group.
She described a finance organisation at a crossroads in 2022. "We had to remediate for an end of life, end of support problem and we also had a unique business problem we needed to solve for." Since then the group has migrated group reporting and the ledgers and has completed two successful month-ends.
On lessons learned, she said one of the most significant ones was the complexity of Unicode conversion — in hindsight it should have started much earlier, it proved more complex than anticipated and became a critical dependency, and investing in foundational technical readiness would have significantly reduced downstream project risk.
Two additional points stood out. The first was her insistence on sequencing. "I am a great believer in architecture: that builds a roadmap about how you want to transform. You can unpack the roadmap and incrementally deliver against it." The second was where she now considers the industry to be. In her view, AI is no longer just around experimentation and production use cases; organisations are now going beyond that to look at enterprise scaling casesfor real business value."
On innovation she was equally clear that teams need room to experiment, but inside a controlled environment, partnering with vendors and keeping governance and guardrails in place. For anyone running a regulated institution, that balance between complexity on one side and platform on the other will be a familiar tension. She also said change management matters as much as the technology implementation itself and it's about keeping people calm during turbulence.
Stanley Dube, Head of Solution Advisory at SAP, gave the day its most quoted phrase. He described the "enterprise paradox": the expectation of instant enterprise AI colliding with the operational reality of complexity, manual processes and human behaviour.
He also gave it its most memorable staging, sharing the platform with a robot co-presenter in a tracksuit carrying Joule branding. "I have a confession to make, I have been afraid of AI," he said. "The fear I have had is that AI is going to take my job." The demonstration that followed worked through how organisations redesign jobs and processes, hold governance and build advantage.
We had the good fortune of a well-positioned stand, which served coffee in the morning, drinks in the afternoon and became the gathering point for prizes at the close. That gave us a full day of unfiltered conversation.
Three themes came up repeatedly.
Is Joule genuinely more than a chatbot? This was the single most common question. People have seen the demo. They want to know what it does on a Tuesday afternoon in a real finance function.
What about those of us still on premise? A lot of the vision assumes cloud. Delegates running on-premise estates wanted to understand which parts of this apply to them now and which parts require a migration they have not yet scheduled.
Data, again. Every conversation eventually arrived at the same place. Nobody disputed that the data foundation determines whether any of this works, which is a notable shift from a few years ago when data was the thing everyone skipped past.
If a client asked us on Monday morning what this all means for them our answer would be simple.
The direction SAP is taking is genuinely powerful and it needs to be weighed against your own context rather than adopted wholesale. Most organisations already hold more capability than they are using. Getting a clear picture of what you have entitlement to, what you are actually running and what is realistically possible from there is the most sensible first move on any AI journey.
That is unglamorous advice. It is also, on the evidence of a day spent listening to people who have done this, the advice that separates the organisations making progress from the ones still running proofs of concept.
Our thanks to SAP Africa and the African SAP User Group for a well-run day and to everyone who stopped by the stand.
