 So I'm David Dickman and I'm responsible for modeling tools at SAP and what we've been finding a lot of people face with these days is an increased complexity. Now we've talked about database systems and sure databases themselves are getting more complex but it's not about this project-centric application-centric I'm building a database for this purpose anymore. What we're finding is people are trying to treat information as a corporate asset. So semantics is a set of web standards that were designed to describe data coming from anywhere on the web and this is a set of standards that's about 10 years old and what we discovered at IBM was that you know this was a way to capture the relationships and meaning across data not only on the web but across enterprises as well. And so when you take these fundamental concepts of linking data together and using a graph model of how to enterprise data you can now easily link together sources of structured data like your relational databases and data warehouses together with spreadsheet data and even unstructured data coming from the web, social media and your documents and emails within the enterprise. The easiest way to explain what a graph database is I mean most people know about databases with the rows and columns but in the graph database you have nodes and links between nodes. These rows and links are actually unique URLs so you can build up a whole set of databases and as long as you make sure that the nodes and the links have a standard vocabulary you can link all these databases together. So what we've done is we've grown data modeling tools up into information architecture tools so we have things like an enterprise conceptual model. This treats for example the idea of customer one instance of the customer entity for the entire enterprise for all business use cases one concept of party one concept of application or component all these things are single instances that have a super set of everything that's in there. This helps us with master data management with master data governance this also helps us align all those different disparate physical systems and start to reduce the impedance mismatches that integration and federation systems are dealing with today. The semantic models give you a conceptual way for business users and analysts to really understand all this data that's been brought together so they can be really involved in building solutions and applications and driving requirements and ultimately understanding their data assets in a way that with traditional data warehousing technology is they're not able to. Yeah if you try to integrate say data for 400 different databases then each database is a schema and for the people in the company it's one big mess because how do all these schemas relate to each other. So what you can do is you can turn each schema of each database into a little graph do some work to link them together and then you could say how does this element in the bill of material relate through all the other database to maybe an ERP system or to a sales database. So that's one way where instead of going through the databases you go through the meta space as if the meta space were a graph. So we look at modeling today it's not just data modeling anymore information architectures a fusion of application architecture technology architecture process architecture and the business side of data coming together with the technology and that's where things are heading next.