## Notes I saw that there was a large miscommunication between my team and the infrastructure team. My team owned the development and maintenance of a geospatial intelligence platform deployed on Kubernetes, and the infrastructure team oversaw the production Kubernetes environment. The teams agreed that the infrastructure team would make sure that the K8s cluster is functional and ready to install my team's software, including making sure that certain services are available, such as Keycloak, container registry services, and database services. Even though this agreement was in place, the K8s cluster was rarely configured 100% correctly. In addition, our team was wasting days making sure that all services were available before attempting to install our software. Lastly, the only way to have the infrastructure team configure a service correctly, we would have to talk to our manager, our manager would have to talk to the manager of the infrastructure team, and the infrastructure team would plan it in the next sprint. Overall this took weeks before the infrastructure team would configure components correctly, and this resulted in thousands of hours wasted over many production deployments. In order to fix this issue, I created an automated cluster testing platform using Python and Docker. This testing platform, once installed via Helm, provided my team with a user interface that checks the required systems, and provides a PDF printout of the services not configured correctly. With my automated testing platform, my team could learn shortcomings of k8s clusters and alert the infrastructure team within minutes instead of having to manually check every service on the platform.