The waste management platform runs on an Azure cloud, where a fully automated CI pipeline using GitLab has been set up. From the initial code through various testing procedures all the way to the production system, a fully functional system can be generated virtually at the push of a button. Without this fully automated process, going live within such a short timeframe would not have been possible. New features were thoroughly tested and approved several times a day. As soon as a feature was ready, it went into production. To ensure requirements such as high availability and scalability, a monolithic structure is no longer suitable today. Therefore, HyPlus relies on a future-proof architectural concept using microservices. Here, individual processes—such as ordering a container—are defined as services. Each of these services is housed in a Docker container. A Kubernetes cluster manages all Docker containers. If a container is unreachable, the cluster automatically starts a new instance. Test-driven development with over 80 percent test coverage and continuous expansion of test phases up to UI automation, as well as technically documented tests and risk-based UI automation tests, ensured high availability. In addition, services can be added and shut down at any time, ensuring high flexibility and scalability. The architecture is designed so that senders and receivers no longer need to know each other. The individual services communicate via Apache Kafka, which reduces complexity. Furthermore, the decoupling of the services allows them to operate at different speeds, making the overall cluster more scalable. The system’s distributed architecture requires appropriate monitoring. Traditional monitoring focuses on files and applications on a fixed server. In the Kubernetes environment, however, the focus is on services running in containers on any host. The containers can be restarted on a different host at any time. That is why HyPlus Solutions uses the Prometheus and Grafana applications in this project. Prometheus is used to store the cluster’s metrics, while Grafana is responsible for visualization and triggering alerts. The applications’ key functions are regularly run automatically to verify continuous availability and functionality.