“Bain’s Enhancing Dashboards with Gen AI — Quicker Service Access And Better Data Analysis”

🔔 Hit SUBSCRIBE button and be the first to get new videos 🎬 Stay tuned on upcoming videos covering various aspects of financial markets and artificial intelligence. Join us on this journey of knowledge and innovation at the BERYL ELITES conference! 🗓️ 6th Annual Beryl Elites Alternative Investment & Data Revolution - Thought Leadership Conference. Immerse yourself in 18 dynamic panels featuring 85 esteemed speakers from top financial, investment, and technology companies. PANEL - AI in Public & Private Markets. Innovation, Adoption, and the Role of Human Analysts: As AI technology reshapes industries, what does early adoption in finance look like? Will it lead to the replacement of human financial analysts? Current landscape, future expectations, and barriers to innovation. Handling AI challenges such as "hallucinations," and how these affect adoption rates and industry trust. Discover intricacies of AI technology adoption in Finance and get the answers from our panel experts featuring: Moderator: Brian Engelbert CPA, CFA CPA, CFA - Managing Director, Kyber Data Science, TD Cowen Speakers: Aakarsh Ramchandani - Chief Strategy Officer, RavenPack Matthew Angrist , Global Head, Buy Side Sales AlphaSense Richard Lichtenstein, Expert Partner & Chief PE Data Officer, Bain & Company Evan Reich - Global Head of Data Strategy & Sourcing, Verition Fund Management LLC Transcript: (00:00) We've actually been doing some really interesting stuff with structured data sets. So at Bain, there's a company called pixis that we bought about five years ago, and they have, I think, about 55 different alternative data sets. Some are more proprietary than others, but then we get them from different places, and they have all the raw data. And so we use this to support our work with investors. We also sell dashboards to companies and other people. One of the things we found is we've added Gen AI on top of these dashboards, and it's made a big difference. So now, you know, the companies that work with us, they can go in there and type in, you know, what's the market share in Dallas, and it'll just, you know, take two minutes, and then boom, here comes a bar chart that shows it over time. And you know, then it's always on, and the next time they come back, it'll show the next the next period and and that's been really a big deal. I mean, we've seen some of the companies we work with be willing to pay more for dashboards that have that built in. And more exciting than that is we've seen the usage data. We've seen that more people at the company are using the Gen AI dashboards because they don't, you know, it used to be there's some nerd and they know that it's market share. Things on page 30. And, you know, so, so if someone wants to know what the market if an executive wants to know the market share, thing is, they call the nerd, and the nerd goes to page 30 and pulls the data and gives it to the executive, and executive shows to the CEO. Now what happens is that the executives just go in there themselves and type it in, right? He doesn't, he doesn't need that. And so we just see a lot more people using it, getting value from it. So that's been really exciting. And so I think, you know, anyone out there who's got one of those, you know, a quantitative data product, should be thinking about, how do we put this natural language interface on it as soon as possible? Because I think it really broadens what you can do with it. And (01:33) is it SQL based? And then you got human in the loop, because you got to check it all, because it's not always accurate. Could you talk a little (01:40) bit about that. Yeah. So the data itself has been cleaned extensively with a combination of machine learning, AI and lots of humans. So you have to start with a very, very pristine data set. If you're going to use this kind of approach. In our case, we're using SQL queries of a snowflake database, but obviously you could do it with Python as well, or some other solution. We have enough confidence in it that we're, we're letting it run. But I think we encourage people, if it's an important query, to check that, check the numbers and, you know, make not, it's a bait, you know, it's marked beta, right? So we're still doing that, but, but I think we found the results are pretty good. I mean, as long as the underlying data is good, it's pretty smart, and the tool is pretty smart, right? If you ask a question, it doesn't quite understand what you mean, it'll come back and say, oh, you know, you said you wanted to know the market here in Dallas. Did you mean this Dallas, or this Dallas or, you know, it'll, it'll kind of ask a few questions. So I think we feel pretty good about what it's able to do. And, I mean, I'm sure a year from now, we'll feel amazing about it, right? [BERYL ELITES WEBSITE] 👉 https://berylelites.com/ Copyright 2025 The Beryl Consulting Group. All Rights Reserved.