SQL Server already handles hybrid data.

We are dealing with more and more data all the time, and in more and more formats. As noted in this piece, companies are receiving data in new formats, from a variety of sources, and the traditional relational database isn't always the best fit for the processing, storage, and management of the information.

However SQL Server isn't just a relational database. It handles unstructured data well, with the Filestream and Filetable enhancements to varbinary columns. We can even read into this unstructured data with iFilters that we install on our instances. In fact, I'll be talking about these filters at DevConnections on Wednesday.

SQL Server has StreamInsight for complex event processing, allowing you to handle very quick streams of data that could overwhelm your OLTP architecture.  If that's not enough, there are connectors that allow you to integrate with other data processing architectures, like Hadoop.

With all the various subsystems like SSIS (ETL), Reporting Services (exports to other formats), Service Broker (messaging), spatial data, and more, I think SQL Server is an amazing hybrid database platform in and of itself, without any extra charges for the various features (like some other platforms). That's not to say you shouldn't use any other platfoms. If they work, use them and integrate with SQL Server, but if you don't have other databases in place, I'm not sure you need to look past SQL Server for many of your needs.