In last years NoSQL experienced a huge upward trend, offering new data models to solve problems where old RDBMS failed. Now people who have chosen NoSQL as an architecture component, realize that a single data model (even when richer that relational), is not enough for average needs. We will discuss why graph databases are at the heart of the multi-model revolution and why we're approaching the end of NoSQL's fragmented ecosystem where customers are forced to use multiple tools in their architectures. Benefits and compromises of this approach along with real world use cases will also be shared