EY has embarked on an ambitious graph enterprise AI and machine learning initiative to uncover fraudulent activities more effectively. They've chosen graph technology for its simplicity in visual constructs, semantic information, and speed up the development process. The cost of computing has fallen precipitously over the last two decades, making graphs a perfect scale-up solution for cloud-based storage. Graphs are popular due to their ability to connect data across multiple domains with processes that rely on relationships and dependencies to uncover patterns. EY is working towards applying AI and machine learning to graph models to assign additional context, inferences or domain knowledge to make better decisions. A knowledge graph is the perfect tool for achieving these goals. Graphs are becoming increasingly popular due to their growing popularity, cost-effectiveness, and ability to drive growth and interest from the community of practitioners. Traditional relational databases will eventually be replaced by graphs as they compete on speed and relevance. Graphs are a natural fit for semantic representation that's easy to develop and understand. They provide a common data fabric that allows pulling up and connecting important data, and through Neo4j APIs, driving applications. The customer 360 use case is one of the most common graph use cases due to its difficulty in enterprise settings. The B2B use case involves master data management challenges. Financial use cases involve identifying rare patterns like money laundering through machine reasoning inference. Graphs are also useful for network optimization and providing additional context through implied relationships. EY aims to implement features such as current AML processes, graph analytics, and deep learning models using Neo4j at scale. Ingesting large data is done by building graph-form tables of nodes and mappings that can be uploaded into Neo4j efficiently. Advanced analytics can be used to make use of Neo4j data, and productionalizing analytics can be easily achieved. To identify whether or not you have a graph problem, consider questions such as getting better customer understanding, mobilizing and syndicating data, business value from existing data, and the next best action for your company.