 Arun, welcome to ODSC India. Thanks, Hamas. Now you're not strange to ODSC having spoken at our previous events, and I want to talk about that in a little while, but do tell us about yourself and do tell us about your talk you're about to give this afternoon. Sure. I had one of the teams of quantitative research in Bloomberg, we work in the quantitative finance discipline, looking at deep domain expertise in asset pricing models, looking at derivatives pricing, for example, or risk models. But obviously, lately infusing that with machine learning and data science. As you know, finance is a goldmine for data. So this afternoon, I'll be talking about use of machine learning and finance, especially with respect to looking at sentiment-based strategies out of the news data and also more general factor rotation strategies for investing. So definitely there's a lot of areas we're exploring within Bloomberg to bring more automation, more data science into financial computing. Yeah, and Bloomberg is a great company. I used to work in finance myself, use the Bloomberg terminal. Lots of very smart people working there, and really a lot of questions I could ask you. But one very hot area of finance has been for a while, the sentiment analysis, right? Because everyone wants to gaze at the sentiment of a stock, of a bond, of the markets. But a very difficult area at the same time, because so much unstructured data there. So tell us a little bit about whatever you can about that area of finance. Yeah, so this is really the area where Bloomberg sort of started his machine learning efforts with the natural language processing of news data. So we do have the benefit of having large amount of in-house experts with our editorial staff and journalistic staff and industry experts who can give us a lot of training data. So we don't have to outsource our training data, corpus generation. So our experts tell us whether a certain story is positive or negative with respect to the company's mentioned that news or the tweet. And then we apply very complex machine learning algorithms on top of that training data to really find out in real time as the stories break whether the story is going to be positive, negative, or neutral for the entity mentioned in that news. Yeah, very exciting stuff to work on. But Bloomberg, more so than any other finance company in the planet, you've got a pulse into all the asset management and the quantum firms. Like how big is the interest in data science in AI for finance right now that you guys see from your perspective? Yeah, so as you said, Bloomberg is kind of like central nervous system of the fintech world. And the interest in data science is really huge. I mean, a lot of the processes need automation just to begin with processing of news, generation of alerts. We even do generation of news from data-driven signals. So we are looking at all kinds of multi-factor data coming out in the markets whether it's macro indicators or market indicators. And we alert our training community or our users to if there's something interesting going on in the markets. For investment as well, trying to build training signals from newer and non-traditional and alternative data sets. We're looking at, for example, credit card transactions or satellite image data which can tell us advanced signals on how a company might perform. So all of these areas need heavy amount of data science. No, that's very interesting. And you mentioned something earlier. You talked about factor rotation, right? Which is very much an institutional side thing like when to figure out, when to get out of retail stocks or energy stocks or into something like that. Pretty complex concept because that's fraught with risk. And so machine learning has been used for that as well. That's very interesting. Specifically, we're using it for what we call style rotation where you look at some of the investment factors which have been popular over the years called the style factors, looking at value or momentum or profitability. And you'll see that these factors don't have a persistent performance over time even though they are long-term risk premium factors but they will not perform consistently every single period. So if you can link observed market conditions to performance of these factors and then essentially build a prediction model as to which factor will be the winner next period, then you could sort of rotate between factors and build a dynamic multifactor strategy to take advantage of that. Very interesting. And PWC, I know last year released a report called Season the Price and they showed the adoption of AI across different industries and stay predicted between the next three and seven years. There's going to be a 100% adoption rate for AI in finance. The retail institutional side, do you agree with that sentiment? Not quite. I agree with 100%. I think AI in finance, I think we definitely see it making a huge impact in terms of sort of more consumer finance whether it's to sort of look at predicting defaults of consumers or for credit cards or loan data or a lot of automation. But in terms of investment, I think building better investment strategy, I think as we know finance has a lot of sort of inherent noise in the data. So to really build it robustly and provably is a very complex process. So I think AI will definitely help in that process but I think a lot of the onus is on the data scientists and the users to really make sure they are using proper science as well. Oh, of course. But there's also a lot of excitement in specific areas and new areas. Like if you think Robo Advising, right? Which has got positive and negative press around us. I'm giving advice to people via AI. But if you think of countries like India which a lot of people don't have access to financial advisors, for example, like huge potential around that. Any thoughts on the Robo Advisor industry and was it Kencho out of Boston? Yes. Is it very big in that space? Yeah, Kencho was acquired, I think with S&P, I believe. Yes, I think those are like sort of entry level ideas in terms of like sort of doing simple data driven sort of strategies looking at essentially pattern analysis of what happened in the past and similar conditions were there in terms of the market as of today. Then you can sort of study patterns in history and build predictions, things like that. I believe a lot of that has potential is sort of simpler data science maybe, but it's very important. But as we all know, more and more people start using similar methods, the real alpha or premium goes away. So you always have to look for newer data sources, newer algorithms, so that search is like ever continuing. Well, I ruined AI and finances. We could speak for hours, a really interesting topic, but thank you so much for being here at ODSE in the end. Looking forward to hearing your talk this afternoon. Thanks, Seamus. Thank you.