Cosmos DB is a globally distributed, multi-model, Serverless, NoSQL database solution that runs on Microsoft Azure. With guaranteed SLAs, various consistency models, and support for multiple APIs, Cosmos DB can have many advantages over common relational database solutions. However, the shift to NoSQL in addition to the numerous configuration options available in Cosmos DB can be a challenge for traditional relational database developers. In this talk, we will take an existing application built on a traditional relational database and update the solution to take advantage of Cosmos DB. Along the way, we will have many decisions to make including which API we should use, how best to model our data, which consistency model to use, and how our data should be indexed, partitioned, and organized. By the end of the talk you should have familiarity with the decisions you will need to make to successfully implement your own solutions on Cosmos DB.