This text explores the application of Neo4j, a graph database, for fraud detection using Entity Resolution (ER) and Weakly Connected Components (WCC) algorithms. ER is used to resolve groups of individuals behind sets of user accounts based on shared credit card information or devices connected to less than 10 total accounts. The WCC algorithm is then applied to the resolved relationships to partition well-defined communities in a scalable manner. These communities are used to label all users in them as fraud risk if even one flagged account is present, significantly increasing the number of identified fraud users from 87.5%. The analysis also shows an improvement in card and device discrimination among fraud risk accounts.