 This is our team. We have a domain expertise in both investments and technology and ESG and sustainability. Sophie and Jonathan, Eve and I worked together as colleagues at Axe Investment Managers at the Rosenberg Equity Division, which is the quantitative equity group where we dealt with a lot of data, big data. We built models and quantitative tools to analyze companies and build portfolios. We, for a lot of years, fully integrated ESG into ESG dimensions, into our portfolios, into our assessments of companies from the early on. So we never had a non-ESG version and an ESG version. We always felt that those things were material to company analysis. So we've been working together for a while from the investment asset management side. And we've now put our minds together and our skills together and our domain knowledge together to attack the vast amount of information that's out there to help us understand companies and the way they're working through issues of sustainability and how they're being perceived out there. I mean, I think as you've heard throughout this conference, the big challenge here, there's no shortage of data and that data is just growing exponentially. The number of us that are attacking the analysis of that data isn't growing quite as fast. So there's what we call an intelligence gap here. One of the goals of this conference has certainly been to find solutions and tools and ways to shrink that intelligence gap. And another way of doing that is another one of the goals of this effort, the integrate effort, is to build. And what we heard a lot about during the conference is to find ways to make the data that is reported more structured, more consistent, more transparent. What we're working on, though, is another subset of that data, which is a very large subset, which is very unstructured data. So no matter what we do on improving the structured data, there's still a huge component that's unstructured. Some estimates put that data as about 80% of all relevant information. And it mostly comes in this format of documents, news, social media posts, customer reviews, employee reviews, NGO reports, industry research. All this unstructured text and language data is really what we're focused on and what we're trying to understand. What we have to do when we're faced with that much information is use machines to help us with that effort. So artificial intelligence and particularly national language processing helps one underneath it to identify trends to look for emerging issues or crises or companies, reputational risks that might be being sourced or being identified through various news and documents out there. And also, as we heard yesterday, too, about the importance of both reputation and purpose to improve your brand image, you can also pick up signals and information on a company's brand image there. And companies can use this information to kind of understand where they're standing and help protect or improve that brand. The way it works in practice is a variety of technological solutions because there's only one way to do all this and that's through harnessing the power of technology. And it starts at the bottom with extracting data, all this unstructured data, the various news sources. We have feeds that get us information from about 37,000 different publications. We get hundreds and process hundreds of thousands of these documents in real time during each day, and then we map those data and that information to companies. That's that first line of processing where we process that data, bring it in, collect it, map it to companies, map it to industries, map it to countries. And then we connected further in the AI engine to events. There are various events that are picked up in the news and also to categories like sustainability categories from the SASB or SGGs or other customized categories. And then deliver it all through a front end, a SAS platform front end, or customized kind of enterprise version APIs and direct data feeds to companies. The idea, though, is to come up with a sense of what's going on from this kind of crowdsource unstructured data that companies can use to assess how they're doing, how their competitors are doing, and what's going on with trends and various dimensions. And importantly, our first application of it is within the sustainability world. Next, turn it to Eve, who's going to show us or get into our tool and kind of demo it. It's called Iceberg. Tensorial is the name of the company, and Iceberg is the name of the first generation tool. I was heartened to hear some references to Iceberg yesterday and the discussion of tangible versus intangible assets in valuation of companies and the valuation of companies and the idea that a lot of the value of companies are in these intangible assets that's not reported on financial statements, not seen above the waterline necessarily. And there's a lot of information when you go to value a company that's kind of below the surface of the water, below, you know, that's not seen like an iceberg. And we're doing the same kind of thing. There's a lot of information out there. What we're trying to do is bring more of that information up above the surface. So that's that's the name. Stop sharing now and let Eve share his screen and take us through the tool. All right. Can you see the screen? Yes. All right. So this is our tool. This is called Iceberg, as Steve mentioned. First of all, what we want to do with the screen is to select a watch list, which is a list of companies that users are interested in. And as you can see from the left, you have filters to select the list of companies you're interested in. And we are a global company, which means also we look at companies all over the world from every country. So you can select a country that you're interested in. And you get the list of companies that you want in that country. For now we're looking at about 100,000 companies. But we can scale that up to potentially a million companies. We can also, if you prefer to set up companies in a given industry. And again, we look at all the industries. So we end industry agnostic. And you will get for a given industry are the companies that belong to these industries. This is what's called the gigs industries. Or you can go by size. And again, we look at the very large company or very small companies. So we also are going to stick to size. And we also have to stick to status. We look at public or private companies. Private companies are especially useful for supply chain. So you can do that in the filters. And for this demonstration. Once you select the companies that you're interested in, you select the button here that says I want to follow that. And for this talk, I'm going to select PepsiCo as one of the companies we're interested in. So once you select that company, maybe later I will show Pfizer to show that this tool is in real time. So the, the tool I'm showing you now, the news that we get. And the articles I can show you now, we're not the same techniques ago, I will not be the same techniques from now, especially with what's going on with Pfizer. But for PepsiCo, once you select a company, you go to the news feed and you say, okay, I want to look at what's going on with PepsiCo. And then within that you see all the news and all the events. We call them events that are related to PepsiCo. And then we have what's called frameworks. We have the business framework, the environment framework and the social framework. And you say, okay, I want to look at the business framework. And I want to see what's going on with PepsiCo and COVID-19. So there's quite a few things going on. And for some reason, lately PepsiCo has been using, we're talking about mentioning an interest in alcohol beverages. So why is that? Well, we can see that there is a relationship here between what, for some reason, between what PepsiCo is trying to do with alcohol beverages and COVID-19. And we detected two events in that direction. They're both, both about one month old. And this is what happens now to explain to you what an event is. It's an NLP, Natural Language Processing Techniques, to in some sense relate in a semantic sense of what these articles are about. So when I click on this first list here, I get the articles related to that topic, which is in this case COVID-19. And you see in here the content of what we call the centroid of this event that shows the news related to PepsiCo and COVID-19. But I can also go to related articles, which is the clusters of these events that are related. And by going through this cluster, I can see why there is a relationship between what PepsiCo is doing with alcohol and COVID-19. So then we can click on any of these real events within this set of articles, and I can enter the details. So that's an example of an investigation we can do on the COVID-19 site for every company in the world. Now, to show you another example, we can go inside the environmental side of things and we can see what's going on with PepsiCo and Energy Management. Well, I find a number of articles here. And we can see, for example, the second list, which is also a set of articles, a cluster of articles, that relate to what PepsiCo is doing with Energy Management. And here I see again the center of this event, but I can see all the articles related to Energy Management. And you can see the list of them, and there is quite a few. And for each of them, you can see the source of these articles. These sources are also global. They come from every country. And each of them may have a different view of the meaning of what PepsiCo is trying to do with Energy Management, which is basically to go to power operation with renewable energy. And you can see the interpretation of what it means within these articles. And this one, for example, is beverage technology and markets. So it's again a professional premium publication, but they give the one point of view about what PepsiCo can do with renewable energy. I would take another example here, go into the social framework, and we can say, well, I want to see what PepsiCo is doing with employee engagement, diversity and inclusion. And again, we'll find a number of articles here, a number of events. And most of them are like two or three weeks old. And I can select this one, for example. And you will see that PepsiCo is about to invest or invest $170 million in Hispanic business to promote diversity. And again, we can go inside of these events and go to related articles. And we can see, for example, this article published by Industrial Baking and Snacks. And we can see that PepsiCo is also not doing this big investment, but is trying to do things that are more local in this case to change values in Chicago's racial and ethnic wealth equality. And that's going on right now. And you can see the details about what PepsiCo is doing. And see also the highlights of these articles and how much PepsiCo is investing. So that's what we detect with the framework. Now what we can do also from this, of course, we can select any of these categories. And we can go into the details of PepsiCo. And we click on this and we can select, again, either by virtue of PepsiCo. And with PepsiCo, what we see here, actually, is the time series of events that happened with PepsiCo over the last month or so for now. But we have data that it's, we have about 14 years of data. So we can see the evolution of these news. We can select our game. We can see basically the statistics of these events. We can see that the number of articles, we can see the changes of the time. And we can see the peaks. So we can go into the time dimension and say I want to zoom in on this. And I want to see what happened in that day. And we can see all the articles related to PepsiCo on that date. I can also enter, for example, a competitor. I can enter Coca-Cola and I can enter COVID-19. And potentially here I will see all the comparisons of what PepsiCo and Coca-Cola are doing for COVID-19 over the last month or any period that we want. So that's the basic functionality that I want to show you. And I'm going to stop sharing and pass the microphone back to Steve. Yeah. You can bring Brad into and can let people ask questions. You can sort of see the tool, the use, the, you know, what we envision it being used for and what we are targeting at is, you know, companies that want to understand their own initiatives along these dimensions and how that's being picked up in the press and reports and documents, how their competitors are doing along those dimensions, what kind of trends along these topics are happening. And then also, you know, maybe they want to look at their supply chain. They can put their own companies in their supply chain, see how they're scoring or see how the world is talking about them through these various sources. So we'll pause at some people asking questions. Okay. So first of all, thank you, Eve and Steve, a great overview. And it's obviously, this is a powerful tool. I mean, it's not humanly possible, obviously, to parse through the amount of media and the types of sources and information that your platform can do at any given day. And as you said, you know, it could change, particularly if you're in a heavy news cycle like a Pfizer. That's probably changing by the moment as more is written about their, you know, their vaccine and the potential impact. I think one of the challenges that you often see when you look at across media broadly is that, you know, certain publications, they like to believe, we like to believe that they're taking perhaps an objective stance and showing all sides of an issue. But we know I think in reality that that's not necessarily the case. And some news sources have a very sort of persistent point of view. And I was wondering if either of you could talk about your platform may take that into consideration or, you know, how that, how do you track a potential bias in a position from a particular source and how that might get reflected in the analysis? Yeah. Yeah, so I would say that for now what we want, you're right that there is probably subjectivity in these publications. Although you can see the few publications I've mentioned, they're very professional. Some of them are very professional publications. Some of them may be not as objective. For now we mostly leave it to the user to decide which publication they're interested in and the publications they're not interested in for any reason. We are thinking of doing some kind of a, if you want some sentiment indicator about the objectivity or subjectivity of sources. But we're being very careful with that because we don't want to add our own bias to existing biases. It's difficult to, to surprise the bias and not to add your own bias. So we've been very careful with that. So we want to give maximum power to the user to decide what is the bias or not. And one thing to help with that power is to be able to see by types of source, maybe categories of types of source, how that information is collecting. So that you can, you can kind of see if there are differences of how different types of sources are, are reporting that. So that's just a matter of categorizing by source, as opposed to excluding or kind of doing a, a warning flag on this source, or is another source. So Steven, so sort of following along that, that line. Obviously you have already built a number of categories that your, your platform already has in place. Are you considering adding other categories from other frameworks and that kind of thing? Could you elaborate a little bit on that? Yeah. So we have fish started out like a lot of folks did with the SASB type frameworks, adding SDG frameworks, just being at this integrate conference. There's probably about seven more frameworks that I now have on the to do, on the to do list. So there's a lot of frameworks. And as Bob mentioned most recently, you can collapse a lot of those frameworks into certain categories. So that whole idea of not overwhelming, but capturing all the nuances of different frameworks and different taxonomies is important. So that, that's, that is definitely work there. The other component that Eve showed at the start was sort of business dimensions. So sort of corporate result dimensions. It's a little different than and needs to be, and part of this conference also is connecting up those worlds with sustainability frameworks with the corporate results dimension. So at the same time we're building out that set too. It's part of the whole NLP AI effort is it's not just searching on key words, but it's sort of intelligently with your domain knowledge. What are the words and the ways those, those frameworks and those goals and topics show up in these text documents and sort of building the way that the engines can kind of look across those documents and find and cluster these things. It's really where the AI comes into play. So in, oh, sorry, go ahead, Eve. Yeah, I would add also that the technology behind it is very flexible. And currently these frameworks can also be industry specific. And we currently working with a transportation, for example, industry. And we're building a framework that that's specific to this industry, what they're interested in, and we can add it to a framework and categories. Right. So in terms of, you know, Stephen, and Eve, you both mentioned that obviously you're scanning. I want to say, I think it's 500,000 or more sources of information a day. I know you mentioned obviously news events and there's a wide variety of things that are passed over those new sources today. Can you kind of give a sense of some of the other sources like social media and things that you might be, that your engines might be scraping or the other kinds of sentiment that you're measuring in the market? Sure. Eve, do you want to talk about that? Yeah, we started with a major source, I would say a new source aggregator. That's called Navia. And we're looking at about 27,000 different types of sources of publications. We consider that these sources are premium. They give you good information. They're better than the social media in a way. They are more, they go deeper into any topic, but we will, we are in the process of integrated social media and RSS from Google. So that will also come up. So we're increasing our sources as we speak. It's being tested right now. What Eve called RSS from Google is sort of the Google universe of documents, which is a lot, and in multiple languages. So that's the other work, which is to map to languages. Right now we can do all what Eve showed in English can be done in French too, because half our team is in Paris, half our team is in the Bay Area in San Francisco. So those are the two languages at first, but the underlying sources now are in multi languages. And then who would you say are your, to me, I see a couple of different potential users in companies for your tool, but I wanted to kind of get a sense from you. Is it targeting a very specific type of user or give us a sense of who your audiences are. Yeah, I think on the sustainability dimensions, the sort of the CSO, hopefully pushing that up to the CFO, the way you guys have been, we've all been talking about the last three days, people within that sustainability efforts department or group that are trying to show how these initiatives that they're doing are being reflected in the outside world there. So picking up some of that information. Look, we're doing what we just did is resonating, it's showing up in these various sources, or here's what our competitors are doing on these dimensions. Maybe we're lagging behind on this certain trend. So that kind of in the corporate world, I would say communications firms also sort of PR firms that work for these corporations that want to understand how is my company being perceived out there. Can I kind of get a set of news doc in documents that are related to various topics for this company? Certainly kind of regulators and folks that want to monitor companies, investors too. I know when we were managing money at AXA, we often wanted to have stories to tell about the companies that were in our portfolios, good stories on the SD dimensions, or we wanted to know about bad stories that were coming up as well. So investors and then the asset owners who want another dimension to kind of bring their portfolios score, if you will, to light in a way that's not just the metrics that we've been talking about a lot in the last couple of days, but also in a more sort of narrative driven version.