Fireside Chat with Google, Capgemini, and TigerGraph
- Blog >
- Fireside Chat with Google, Capgemini, and TigerGraph
This is an abbreviated version of a conversation between Philip Moyer, Vice President, Strategic Industries at Google Cloud, Sankar Krishnan, Executive Vice President, Industry Head, Banking & Capital Markets at Capgemini, Michael Shaler, Vice President of Partnerships at TigerGraph, and Chris Preimesberger, Contributing Editor at ZDNet during the Fall 2021 Graph + AI Fall Summit conference.
View the full Fireside Chat recording.
Chris Preimesberger: I’m Chris Preimesberger, the contributing editor at ZDnet, and your host for this fireside chat. Philip, let’s begin with you. Can you describe a few new technologies and solutions that you are adopting at various levels of Google Cloud and how they’re being used?
Philip Moyer, Vice President, Strategic Industries at Google Cloud: Yes. I would say artificial intelligence is a very significant focus for us. I would break down the sort of artificial intelligence technologies into a couple of areas. One is Doc AI, which is a really important area for us in terms of using natural language processing and extraction of information. The other is in the area that we call Contact Center AI where we use learning models to be able to get more intelligent to serve in areas like contact centers. And the third area is what I would call Machine Learning Operations.
Also, we just announced a really significant area of technology, it’s called Vertex, that brings together the operational challenges and smooths over the ability for machine learning experts to be able to do their role. I’d say those three areas are probably the most significant areas for us right now.
Chris: Thanks, Philip. Sankar, what are some cool next-gen technologies that Capgemini is recommending to its clients in the financial services market? And what are some of the key use cases? Can you touch on those?
Sankar Krishnan, Executive Vice President, Industry Head, Banking & Capital Markets at Capgemini: Absolutely. First and foremost is this whole concept of the way the verticals have been configured, it’s very old. We look at ourselves as financial services, technology, telecom, and retail. But what your future customer is really saying is, “I want to run my day with better technology.” So first and foremost, what about those technologies that help converge and drive businesses across? That is the first thing.
The second thing is there is a big word out there, but a lot of work to do in the form of hyper-personalization where not two people and individuals are the same. Third, is this whole talk about the metaverse and what should be done today to approach it, especially as Phillip mentioned about AI, machine learning, and a lot of the conversation today around it.
And honestly, looking at Jay’s presentation this morning, I thought the folks at Squid Game were a graph customer, because when the villain comes in and the hero thinks “Here’s why I’m here,” the guy says, “What do you know about me?” And when he goes, “Here is your age, he has a daughter and a network of loans”, it sounded like a graph. We don’t do that. Record. Those are some of the things out there. But more importantly, I think the reason you see I’m cloud-agnostic. So I mean, every type of cloud and where it is growing, and it’s good to see with graph we have a new partnership that has started on the regulatory side, but we are looking at taking it into other areas as well.
Chris: There’s a graph for everything, even for Squid Games. I love it. Phillip, back to you. How is Google Cloud benefiting from graph and AI today? And how is the GCP differentiating itself from customers?
Philip: I would say the technologies that I’d mentioned, for probably first and foremost, I get most excited around Doc AI technologies. And one of the most important things that we do so first of all, we’re benefiting from basically doing natural language processing and 26 languages. Most AI to date is really glorified optical character recognition. And what we’re starting to do, we’ve had some significant breakthroughs and a scenario that’s called, “Bert,” inside of our natural language processing, that allows us to understand context, allows us to understand sentences, and it allows us to actually bridge between sections of documents.
Now, the technology we use with document AI, first of all, there’s a lot of work just being done to be able to maintain fidelity of a document. Like if you take a municipal bond, as an example, I should say, a bond, that might be a Thailand bond, you can actually use the natural language processing to maintain the fidelity of that document. And it’s super, super important to be able to maintain fidelity and context up and down inside of the document. But what we do is we actually branch over into the Google Knowledge Graph.
And so we’re starting to be able to provide access into the Google Knowledge Graph. So if I’m parsing something like a commercial mortgage-backed security, and I want to be able to determine the rent rolls, whether or not those rent rolls are or are not fraudulent, I can check to see if the individuals I can quickly hit the Google Knowledge Graph and say, has this company ever actually been using an email address? Does it actually have an address? Is it live on the web, that inside of anti-money laundering? That’s a huge, huge issue, you’ve got a lot of people that are sitting there googling things.
With our Knowledge Graph, we’re now starting to use our knowledge graph to be able to join together with customers’ knowledge graphs to create their own unique knowledge graph and be able to parse the information more effectively and faster than a human could do it on their own. Many of you were in the breakout session on financial crimes, that was a very popular session today.
Chris: Sankar, what are some of the ways that Capgemini’s clients are using graph and AI to fight financial crimes? And what are some of the economic benefits that Capgemini is forecasting for clients in that area?
Sankar: Yeah, first and foremost, is in terms of fraud prevention. So I think we take a lot of the capital markets, and you touched about some of that, Philip. You take a lot of data relating to capital markets, activities, swap activity, and so on, and then you’re trying to look for patterns, where banks have paid fines before. Then, you’re working back and saying, what caused some of those? And then now we are taking all your customers with a similar profile, putting them on graph to see if there is a similar pattern.
The second interesting thing is that in terms of segmentation, if you look at wealth management and asset management, as an industry, a lot of things are segmented the old ways in terms of high net worth, mass affluent and the digital natives are today doing things very differently. So there is a need to look at these copious petabytes of data and kind of come up with new ways of segmentation, right? And this is true at an entry level when someone is just beginning to open a Robinhood account. It’s equally true when someone is a Boomer handing off that wealth to the next generation. So that is a good example.
A lot of wealth management goals have been made using, how does this consumer act? How can we look at him on a graph and then say, here is the way he accumulates? But now with Baby Boomers, you have to do the inverse, which shows the way he accumulates because each of us has to withdraw $100,000. None of our financial advisers are telling us the smartest, tax projected, adjusted way to do so. And initiatives like the graph can help us get there. And also I spoke about how all the verticals are coming together in an IoT setting.
So for this unique graph to kind of say, “How was Sankar?”, the telecom customer connected to Sankar, the financial service customer connected to Sankar, the retail customer, and so on? And then how do you develop a buying behavior algorithm on top of that, so that you are able to service customers very differently? So, we think that this whole thing can shave 30% costs over a two-year period, depending on the way you deploy that.
Chris: That’s impressive, 30%. Michael, graph and AI are new for many people. As we discussed here in this conference, how do you describe derisking, this new technology for those of us who are really interested in risk?
Michael Shaler, Vice President of Partnerships at TigerGraph: It’s a softball because we work with these partners; that eliminates a lot of the risks, but we also make it easy for customers to get started and in very nominal ways. We have a free tier that we run on all three clouds, we actually have the ability then also add solutions on top of that.
I think that you know when we were talking about the AML in the Financial Crimes breakout. We talked about AML in the context of, how do you build a solution around that? What are the data sources needed? What is the strategy, what is your cloud migration strategy? And so we’re able to kind of stitch that together in a way that reduces risk and accelerates time to market. So we actually have a very strong story there together with our partners.
Chris: Okay. How can a customer, Michael, get started the most easily? And how is cloud a stepping stone for getting started?
Michael: Well, all your data is on the cloud anyway. So what Phil is describing in terms of Doc AI and other capabilities of the cloud is very powerful. And one of the things we’re seeing is that it’s easy to do proofs of value on the cloud. And then from there, customers can choose to stay on it, or they can actually kind of move it into on-prem. That’s all about the kind of governance. And the other risk elements of that summary is that they can get started on our free tier, leverage our toolkits, and then get going fast.
Chris: Thank you, Michael. As we draw to a close, I want to thank everybody for joining us today. I know I’ve learned an awful lot. I’ll be taking home a lot of takeaways. And I thank all of our speakers and our presenters here today who brought such great perspectives to this topic, which is only going to grow in importance as time goes on. So thank you all, you should all give yourselves a round of applause.
Spring 2022 Graph + AI Summit
The Spring 2022 Graph + AI Summit is just over a month away, and registration is open. Don’t miss out on the industry’s only open conference dedicated to democratizing and accelerating analytics, AI, and machine learning with graph algorithms. Register for free today!