Thread regarding SAP layoffs

SAP needs Indians on the board. Then SAP will win the AI battle big time.

@a4 Trust your board, eh?

@mb But AI is dependant on it, and I specifically talked about hybrid cloud, the core erp engine runs on customer managed infrastructure, but by modernising abap so that its not run in the database but as truly standalone layers communicating via apis you can pull the compute layers out of the core product and run them in the cloud next to the ai layers that need specific hardware etc. The reason we can't do this is tied up in the fact that s4 is a monolithic design due to ABAP. But due to ABAPs simplicity that could have had a redesigned execution engine that allowed for shared compute, isolated data model that clients want.

@k4 Ever heard The Innovator's Dilemma? Cloud native ERP was tried unsuccessfully if you know the history. The current solution is actually more acceptable to the existing customer base. Modernization of the stack is also being tried but having very limited success. The issue is not with not having a cloud ERP solution but with not having more innovative features that customers need/want and easy to consume. The fact is cloud solutions are not necessarily better than on-premise and not every customer needs cloud-based solution. For some on-premise solutions simply work better and easier to meet business requirements.

@jw The problem is the architecture decision to not build a cloud native erp but instead package up an on prem one with some fancy database layer to look like one. So were now in the situation where we can't offer hybrid cloud solutions easily as all the data is tied up in one big giant monolith. And this all comes down to having developed most of SAP on a 4gl stuck in the 1980's and not modernizing that stack. We are where we are but it means SAP core is dragging down the LOB's that are actually already modernized stacks.

@OP new API policy is telling, we're starting to work out we're just a d-mb database that the next gen enterprise tools like Snowflake, Databricks, Palantir AIP will connect to and become the actual value layer.

Another problem is we've already sold heaps of AI, but don't have a clear path to deliver. Worse, we do demos and show inbuilt AI, and when it gets to the architecture, we have to talk around in circles as theres so many things needed to get the AI to work, or worse, customer is on private cloud so might only get the capability in a few years.

@dx People said the same about electricity, automobile, the personal computer, the internet, the mobile phone,... If you think AI is not a revolution, potentially a bigger one than any other in human history, thing again.

SAP strategy makes sense and could be a winner if executed correctly and demonstrated well in SAPPHIRE.

If SAP develop agents that can execute tasks on any SAP module/product so the customer only needs to automate/orchestrate using N8N style, it could be a huge win.
Basically showing the customers that yes they can build their own agents but they will waste time and money and the result won't be as good as the (freely) provided SAP agents.

In the opposite, if SAP agents and Joule aren't as good as the competition can provide over SAP modules/data then it's the beginning of the end.

@d7 or we put the data in one.place and the customer connect a third party AI solution to do the analysis

@bz HR Droid is too soft :D

LLMs are a losing game. I cannot see how any provider can generate enough cashflow to pay for the current resources they need. Let alone paying back all the fantastic sums that were borrowed and sunk into GPU-based compute infrastructure.

We only use this and embed it in our products. Waaaaay cheaper, and if the costs explode in general, well, we can just pass them on to our customers, and they decide whether it is worth it for them.

We have far less skin in the game than OpenAI, Grok, Meta, Google, Anthropic or Oracle. If AI works out somehow, we profit. If it is a disaster, it is not our disaster.

To me it looks as if SAP wins the AI battle, no matter how it ends.

SAP’s strategy at sapphire is to “flood the zone with chatbots” and ask customers to put all their data in a single repository. This is a disaster waiting to happen because either we get hacked, or a third party provider or partner gets hacked or CK simply sells this data to his real bosses at Palantir.

SAP is winning?
Then why are third party support companies seeing exponential growth from SAP customers who are already happy with their On Prem systems (for which they already purchased perpetual licenses) and just want support for their enterprise systems while their business processes are working well already?
Don’t be so gullible.
SAP isn’t winning. CK, HR Droid and the Toothbrushing expert are the ones that are winning (at our expense) while they get paid handsomely to pick over the bones of what was once a leading company.

#1 complaint from customers for Rise? It’s too slow why are we paying for this? If our core product is flopping why would a customer trust our chatbot?

@a4 - I am guessing you do not talk to customers - SAP is not winning. SAP is losing so badly that times will get desperate, and soon.

@a4 lol

https://dsag.de/wp-content/uploads/2026/02/DSAG-Investment-Report-2026_AI-1024x576.png

You are wrong. SAP is winning. Oracle is losing: they laid off tens of thousands to sink more money into hardware that will become obsolete once the bubble bursts.

SAP has the right strategy. It also has the right people on the board. Trust them, they know what they are doing.

Once the bubble bursts, only reasonable AI applications will be profitable. And SAP has them.