Date: 16.10.2025 (2025-10-16)
David Dür (16.10.2025)
Questions: • Arbeitsalltag • Typical project • Automation? Which tools? Challenges? etc.
Notes: • Legacy systems: the problem is super large • Even with new systems Excel is sent back and forth -> user is a problem -> they do not know what the system can do -> users are not trained -> one trained user who knows the system -> they leave and then there are a lot of people who dont know the system • Data silos is a big problem: they do not exchange data tools between departments • Slowly • Change management is not accepted -> not good enough trained -> do not know why new tool is required • They mainly do new implementation or transfer to new system • Industries: • Telco, mobility: relatively good -> data in the cloud • Regulated sectors: Banks, healthcare (gut feeling) • Typical project: • Check status quo -> which process steps in the value chain, which departments -> vendor selection process -> tender (send them a requirement list) -> long list (10 vendors or sth) -> invitation to pitch • Often custom code is required -> risk that it takes long -> they normally not take that vendor • Middleware / integration is a big problem • IT landscape is very fragmented: Deutsche Bahn has 150 applications - most of them not connected • Problem is not a technical problem -> it is important to understand that there is value -> there are so many KPIs that can be tracked - which ones matter most? • Building an integration is super expensive • Extending a tool is ok • Data privacy is a problem -> mirror privacy across tools • Example: implemented 4 tools; middleware was a fail -> they wanted to have a single source of truth -> there were logics in there -> Dienstleister sind nicht gut (good people go to Google etc.) • Check the AI Act
Feedback: • Make example more realistic -> clear that it creates value • Use Datev as a client and distribution channel • Challenges: who implements the solution. ICP: • SME: might be scared that they are locked in; could make sense to offer a service; some SMEs have enough IT to use it • Large companies • Concrete examples: not only in legacy systems - also bei software where APIs are bad • Infrastructure area: Glasfaser in AUT -> it is not replacable -> there is only one solution available • Note: Their incentives are not aligned to make the legacy SW better -> they want to introduce new software - thats what they are paid for as a consultancy • During cloud transformation -> first microservices are put into cloud -> how do you ensure that Cloud and legacy talk to each other without API • Easiest: put legacy into cloud -> but same problems -> most people put parts into the cloud -> what will i anyways not change -> put it into cloud • Then there are topics where you want to be innovative -> in the cloud via microservices -> how do you connect with Cloud • Typical customer: • Infrastructure project: not a single IT guy that can use the technology • Telco: always big • -> assumption: often there are no inhouse IT guys who can do things - just admins • Pitch to a customer: • Most important: list 4 to 5 examples that are concrete -> example calculation to monetise the value