SAP Technology Versus Business
![](https://e3mag.com/wp-content/uploads/2024/01/Editorial.jpg)
![](https://e3mag.com/wp-content/uploads/2023/12/pmf_RGB_400x400_2-150x150.png)
SAP's love of technology, however, is undeniable. The ERP group has enthusiastically embraced the development of an in-memory computing database. Of course, the initiative came from SAP patriarch Professor Hasso Plattner, who had all the resources available for the creation of the database at his university institute in Potsdam. SAP was happy to take up the database challenge. And so Hana was born.
SQL database Hana
From SAP's point of view, the SQL database Hana is more than just a competitor to Oracle, IBM, and Microsoft. Hana is meant to be a database platform that offers features that go far beyond a relational database. This approach may have been in Hasso Plattner's mind from the very beginning, but the claim to be more than just an SQL machine came late, perhaps too late.
In addition to the SQL engine, Hana already had the option of graph computing in its early years. Graph theory is a branch of computer science. Graphs can represent relationships between objects and are relatively easy to handle with matrix computations. The Hana graph engine has not become very popular, and there are few known practical applications.
Now, SAP Chief Technology Officer Jürgen Müller has surprised us with a Hana vector engine that will support and drive future LLM (large-language models) in generative AI. Graphs and vectors are two very different approaches. However, they can probably be orchestrated on a database platform like Hana.
Vectors span space and are usually multidimensional. What inevitably contradicts the normal concept of space are vectors of a higher dimension, sometimes over a hundred. For humans, the third dimension is usually the end of the line, perhaps with time as the fourth dimension. However, vectors in multidimensional space can be used very easily and well, for example, to find similarities between object
Technical innovations and SAP
Professor Hasso Plattner and his former PhD student Jürgen Müller may find a lot of mathematics and computer science fascinating. However, the added value they bring for the majority of SAP customers is still limited. This raises the question of whether SAP's purpose should really be in technical innovation, or whether it would be better to follow the adage "stick to what you know" and instead focus on driving business innovation.
With all due respect to SAP's Chief Technology Officer, Jürgen Müller, it has to be said that others can do better! Last year, Jürgen Müller introduced a copilot for building software on the SAP BTP (Business Technology Platform) at SAP TechEd 2023 in Bangalore, India. When asked, Müller explained that the co-pilot was JavaScript-capable, which can be helpful in an SAP ERP environment.
Asked again about SAP's own programming language, Abap, Jürgen Müller had to admit that the copilot is still learning. Even six months later, there still seems to be no progress at SAP with regards to an Abap copilot. On the fringes of a BTP conference, however, it was revealed that an experimental Abap copilot already exists in the Microsoft labs. This is not a surprising development! Microsoft is one of the main sponsors of OpenAI, the operator of ChatGPT, and thus one of the world leaders in generative AI, which includes copilots in the broadest sense.
Stick to what you know
If Microsoft really does have an Abap copilot, I don't see this as a disadvantage for SAP and the SAP community—maybe more as a personal defeat for SAP's Chief Technology Officer Jürgen Müller. Why is that? SAP does not need to be a technology company to be successful worldwide. The business knowledge gained from 50 years of ERP development is unique. The world's problems in logistics, finance, controlling, and supply chain planning are enormous. SAP has every opportunity to create a sustainable unique selling point. So why cloud computing, in-memory computing, copilots, and generative AI?
Stick to what you know, is what I want to shout at SAP. There are enough challenges for SAP in the areas of corporate and business organizational structures and processes. Where is the business and organizational value of Cloud ALM, Signavio, and LeanIX? SAP could be on the right track if SAP CEO Christian Klein would talk more about enterprise architects, process mining, and composable ERP and less about cloud computing and AI.
1 comment
Dieter Richter
Microsoft berichtet schon letztes Jahr öffentlich über den internen Einsatz des Abap-Co-Piloten:
https://youtu.be/nixkfXt1nEs?si=yh8A8GjN84nOaz2Y
Von SAP Joule sieht man immer noch nur Videos und das erste SAP Joule Release scheint ein Conversational Search Anwendungsfall auf Basis des SAP Help Portal zu werden.
Ähnliche Anwendungsfälle haben viele deutsche Unternehmen schon lange selbst und innerhalb weniger Monate implementiert:
https://www.sueddeutsche.de/wirtschaft/kuenstliche-intelligenz-otto-dm-chatgpt-mitarbeiter-gruende-1.6309361
Hat sich dafür die lange Wartezeit gelohnt und wie viele Jahre hinkt SAP Joule nun dem Co-Piloten Markt hinterher?