Daimler hat eine neue Cloud-Plattform mit Microsoft gestartet, um datengetriebene Innovation zu ermöglichen. Sehr spannend!
Keine News mehr verpassen! Jetzt den Newsletter abonnieren .
TRANSFORM: What challenges did you have with your on-premise data platform?
VETTER: We had a monolithic environment and limited capacity. The requests and demands for service from Daimler’s business units were so massive that we were not able to scale the calculation power to what we needed. We had our units competing for calculation resources and we had to schedule and plan who was calculating when. But with Azure, the big advantage is we scale up, we compute, we pay, we scale down.
TRANSFORM: How did Azure Key Vault help?
VETTER: Azure Key Vault was the lever for us to move into the cloud. The biggest challenge for us internally is that we process confidential data. We don’t want this data to leak anywhere. But with “bring your own key” in Azure Key Vault, we are in control of the data and encryption material. Nobody but us can use the data. Combined with services like Azure Active Directory, it gives us all the data protection and security we need to make sure everything is to our highest standards of security.
TRANSFORM: How long did it take you to develop eXtollo?
VETTER: We launched the idea of eXtollo in workshops with Microsoft in January 2018. Then we went live for Europe in April. It was a three-month exercise and a lot of the time we spent validating concepts. We went live in the U.S. in October and later in Asia. So, we have in nine months of almost-global coverage. It was lightning-fast. Cloud is really bringing us the speed and the flexibility to do that.
TRANSFORM: What are some use cases for eXtollo?
VETTER: We do a lot of forecasting cases. In the past, it took days to calculate forecasts in the finance or production areas, which the algorithms can now do in minutes and seconds. We also do driving behavior analytics and forecasting with AI on what a customer potentially wants to buy and that gives us a better portfolio of sales.
The best use case scenario is error code forecasting for the vehicle. When you take your car to the workshop for a repair, the mechanic can download an error code log from the car and immediately see how to solve the problem. The computer program is based on a machine-learning algorithm that analyzes historical, diagnostic data of cars to give targeted suggestions for faster, better service.
The cloud really enables the potential of AI for all levels of the organization. That’s one of the advantages of Azure – we have all the tools we need, whether we use AI algorithms or normal advanced analytics algorithms.