Cloud Foundry hackathon win with a new AI Integration Service

At the European Cloud Foundry Summit 2019, which took place last quarter in the Netherlands, a joint team of Bertelsmann, SAP, and Arvato Systems employees won the hackathon. With a solution combining Artificial Intelligence (AI) / Machine Learning (ML) and the Cloud Foundry Application Platform, our hackathon participants created a whole new business case, generating real-life benefits for businesses.

After gaining insights on the specific Application PaaS (platform as a service) that is offered by Cloud Foundry, our team attended the yearly European Cloud Foundry Summit. The Cloud Foundry PaaS is very interesting for companies as it is a DevOps-focused solution for cloud-native applications, which was born in the early days of container technology and is still supported and extended by huge enterprises. Also, the PaaS solution can be leveraged in the same way on many different cloud providers without needing to have very specific cloud expertise on all of the offered cloud solutions, e.g. Microsoft Azure, Google Cloud, AWS, or VPC. Thus, developers can create one cloud independent solution, which then can be hosted on any cloud.

Within the Summit, a hackathon was scheduled to extend or improve the PaaS solution in a unique and innovative way. Our team competed with their project named "Cloud Foundry Artificial Intelligence Project Support" focusing on the extension of the platform with a native AI integration, which breaks down AI projects to the basics: The data as the minimum information needed. This AI integration can be seen as a bridge between developers and data scientists. Data scientists with basic IT skills, who do not have the nerves nor the time to spend days and weeks figuring out AI/ML library dependencies are enabled to host their data modeling and prediction jobs in a secure and scalable way. Also, computer scientists with little or no AI or ML domain knowledge will be able to leverage the advantageous Cloud Foundry push experience to efficiently turn data into AI APIs. With this innovative integration, data scientists can focus on their data model optimization and do not need to build individual services, which then need to be configured to work on specific clouds or data center servers. Thus, this integration streamlines timely and tedious work, allowing a faster turnaround with fewer errors based on this prebuilt service while increasing the offering to less development-focused users.  

Maurice Brinkmann is a Cloud Consultant and Architect with Arvato Systems, who currently focuses on API management and platform as a service solution for applications and cognitive bots. Maurice enjoys supporting the business by driving challenging IT projects and diving deep into innovative showcases. His professional background includes diverse roles as DevOps Engineer, Scrum Master, IT Infrastructure Architect and QA Team Lead in different enterprises.