Bechtle has successfully passed an audit, proving our capabilities in delivering end-to-end AI and ML projects from design to implementation and post-deployment. “We’ve worked very hard towards this Microsoft specialisation because our customers need an expert partner to support them roll out artificial intelligence in their organisations,” says Elisabeth Vinek, Business Unit Manager Data & Analytics at Bechtle.
For many of these customers, the critical first step is to identify the right use cases, which differ according to industry and company size. Fortunately, this is relatively easy with Bechtle’s three-part customer workshop.
Focus areas of our workshop series:
- Demystifying artificial intelligence
- Understanding the challenges
- Typical business cases
- Live demo of AI capabilities
- Brainstorming company-specific use cases
- Alignment and prioritisation of ideas
- Demonstration of ideas
- PoC estimate and proof of concept
- Estimates for macro planning
- Conclusions and recommendations
Data protection-compliant ad-hoc queries.
One customer who followed the Bechtle approach is a Swiss hospital, which collaborated with Elisabeth Vinek’s team to define initial use cases. For instance, this involved medical staff queries regarding the administration of medication to specific groups of patients showing similar symptoms—in natural language and in compliance with data protection regulations.
This simple process taps into a large database to deliver ad-hoc information, which would previously only have been feasible with the support of the IT department and never at the speed that is now possible. “Artificial intelligence is democratising access to corporate data,” explains Elisabeth Vinek. A major achievement for a data-driven organisation.
Artificial intelligence is democratising access to corporate data.
Elisabeth Vinek, Business Unit Manager Data & Analytics
Elisabeth Vinek
Technically speaking, the solution is quite simple. The AI itself is confined to the Microsoft Azure Cloud while the hospital’s data, including confidential patient information, will never leave its premises. AI can only access the meta data, not the data itself. “The greatest challenge in these kinds of projects is often integrating all the required systems, making it just as much about software development as in-depth data expertise,” says Elisabeth Vinek.
The importance of a data strategy.
The data expert and her team are currently working on optimising the way data is being visualised for the customer so it’s easier to draw conclusions and identify other business cases.
One critical finding is, the more extensive and complex the project, the more data handling will shift into the spotlight, which is why Bechtle advises its customers on data storage, integration, analytics and strategy. After all, for AI to render optimal results, it needs to stand on a truly solid foundation.
