Yaacov Mutnikas
Analyst · Deutsche Bank. Your line is open
So, just I'll answer - I'll comment in two ways. One, a little bit more color on the data lake itself and two about some of the examples of analytics that those cross border, cross - across different division lines. So just dipping a little bit into the data lake, so Lance mentioned the data lake is capable of processing and is processing today, structured, semi-structured, unstructured data throughout. Number two, it's rooted in a coherent data catalog in data governance culture, where we've got now a data governance machine, how we manage data that's emanating in the lake, but that also creates a structure across the organization, how we manage our data. Number three. We have an approach where we have got a process for the end of the year to hydrate our Data Lake with quality curated data. Number four, the data science, it doesn't address just issues of product innovation, but widely applies to data curation to which we have a process in place. Second of all, I mentioned earlier is that we've got a way of since the data lake contains all the data across the board, across different business lines, we can now build products that we can merge data from, let's say, energy and financial sector through an unified data interface and cataloging capability. And so for the data lake in theory and in practice supports data management, data creation and product innovation. In terms of - just one part of the color to the thing, of course this entire exercise has to be done in the context of managing the cost of all the data management and data curation. And finally in terms of data science specific projects, one of the project that I wanted to mention is for example commodities at sea, where we can understand from our maritime business all movements of oil ships basically in the real time, out of that we can subset all the energy movements, all the oil movements across the world, we know any oil carrying ship where it's coming from, where it's going, at which point in time it's going to various points where these potential challenges, let's say like Strait of Hormuz et cetera. And we can understand and anticipate when the oil is going to reach various commercial centers and what potentially it can influence in terms of the oil value. The other example that I will mention, we're launching in January a product which is our dividend focusing thing which is now supported by advanced machine learning techniques that going from manual effort of focusing 3,000 companies, we can focus 28,000. And finally we've got a cognitive processing for news as it relates to terrorism, as it relates to civil unrest, political unrest, energy events and similar that all machine is now taking on data from roughly 16,000 open source sources to process classified data and help our analysts to opine on issues that are [indiscernible].