In the big data space, a number of important application domains, such as recommendation engines, fraud detection systems, and social networking, increasingly rely on graph data models to represent domain entities and relationships. In this talk, we present main use cases for graph databases and applications, the Apache TinkerPop project, the property graph data model, and the Gremlin graph traversal language. We use Gremlin-Groovy to demonstrate different types of graph traversals, including simple, branching, recursive, declarative, and domain-specific traversals, in the context of a movie cataloguing application that tracks movies, actors, directors, screenwriters, cinematographers, production companies, as well as users and their movie ratings.