The most important thing you have to do in this day and age is to make sure that you are not chasing a specific technology just for the fear of missing out; it has to make sense for business. I always start with the business use case. We start with the problem which we are trying to solve. And then we look for the relevant technology in the realm of either AI, or machine learning, and so on, and not the other way around. Because there is so much of Insurtech, and Fintech, and other new things happening in the industry, it's very easy to get lured by everything. But in the end, you have nothing to show for. So what we have done is that we have found out some specific areas to explore. Let’s take fraud detection for example - we are now able to use a lot of algorithms to be able to detect the fraud beyond the data points which we have. And essentially be able to leverage lots of those hidden data points and features of a personality and behavioral aspects to be able to detect the fraud in a much better manner, which is not purely based on just the financial data points, but also other behavioral data points as well. Second, in our insurance business, we are leveraging this a lot in underwriting as well. Now, underwriting has been very, very painful for many years, and has been traditional as well. And it has a direct impact on customer journeys because every time a new customer on-boards, the underwriting journey becomes so complicated that they get very frustrated during the process. So we are going to leverage a lot of that data, machine learning to constantly train our own underwriting questionnaires. And really see that what is it that makes sense, and how can we simplify underwriting. So once we simplify, the second step comes that can algorithm make the decision as well. I don't think we have reached a stage of algorithmic decision making yet. Where we have reached is to ability to at least take those data points and redefine and simplify the underwriting. Now the next step is for the algorithm to decide.