New AI platform aims to help users invest

Generative AI tools like ChatGPT all the rage right now. Analytics firm Toggle AI is looking to use the technology to help investors. CEO Jan Szilagyi joins Yahoo Finance Live to discuss how AI might impact investor finances.

Video transcript

BRAD SMITH: Generative AI is all the talk these days and the rise of AI tools like ChatGPT. Investors, they're curious to see whether or not chatbots can help them make tactical investment decisions.

Analytics firm TOGGLE AI is looking to be a key player in this upcoming market by training large language models to predict the price action of stocks, bonds, commodities, and even ETFs. The CEO of TOGGLE AI Jan Szilagyi joins us now. Great to have you here with us on set.

JAN SZILAGYI: Great to be here. Thank you for having me.

BRAD SMITH: So for those who may not already know, for what TOGGLE AI does and how you engage with portfolios or potential investors as well, what is the first thing that they need to know about what this type of AI does in terms of providing perhaps recommendations even or some type of guidance in how to navigate a market environment like this?

JAN SZILAGYI: So I think you captured it perfectly in your description in the sense that what we have really done is, we've taken this really incredible technology that allows us to communicate with computers and train it in a lot of analytical methods that allow you to then understand better what's going on in the market. Just like you were doing this morning to really try to get at the narrative of a range of assets. And I think this has become substantially easier with all the progress we've seen in generative AI.

- So with what you all do with your business, with the analytics, what is the direction one should take going into-- like how do you all analyze the market and how do you all give recommendations in terms of sector? I know we can't go into like specific stocks, but let's talk sectors. Where do you-- how do you all direct consumers? How do you take this information-- not consumers, customers-- and give them direction?

JAN SZILAGYI: So the best way to think about this intelligent data analytics platform, that is TOGGLE, is that it sits on a very large mountain of data. And it's constantly giving a number of these assets a daily health check. What's the trading volume like? What's the macro context in which these businesses are operating?

And so it's highlighting, for the users, what they should be paying attention to. Where is the economy slowing potentially going to be really impactful? Which sectors might get hurt? Which ones might benefit, and so on?

The exciting thing about doing that through the technology we have is, you can ask follow-up questions. So instead of the machine just telling you, oh, this is what's interesting. You say, well, but let's say, I change this assumption, what would you tell me then? So it's a lot more of an interactive back and forth than I think people are used to when it comes to data analytics.

BRAD SMITH: Does this outperform, in your testing and advisor, a traditional method that many who are starting up a portfolio or just looking for some type of guidance in their own, either financial planning long-term or just portfolio management strategy would have gone to? How does this outperform that?

JAN SZILAGYI: So very importantly, it doesn't issue buy or sell recommendation. So what it will effectively do is-- we have a lot of advisors who are actually using our platform. It allows them to very quickly do a data check. For example, you're trying to understand what is currently going on with banks? Why are banks, for example, up today versus technology might be down? Is it more earnings driven? Is it positioning driven, and so on?

So it provides context by harnessing this large amount of data and distilling it down to the key things that really matter. But it definitely works in tandem with us as opposed to replacing advisors, as opposed to replacing what all of us do.

- To your point, I mean, you said advisors. But we mean humans.


Exactly. So let me ask you. So what is the upside to how TOGGLE AI operates with regard to? Is it speed? What is your advantage? What's your competitive advantage?

JAN SZILAGYI: Great question. So the main advantage whenever it comes to this much machinery is obviously it can just process a much, much larger amount of information. And so what it really does is it's tracking these different relationships and then highlighting whatever matters.

The reason why I was comparing it earlier to a health check is that there are so many things that you could look at us to. Do you have fever today? Are you coughing, and so on? And same things happen in the market. We would struggle to go through everything every day. Read the Fed's statement, and so on. But computers don't. And so then, they can summarize, in a few bullet points, the three things that you should pay attention to today.

BRAD SMITH: So business-wise for TOGGLE AI, what are some of the key performance indicators that you look to say whether or not TOGGLE AI is a success; margins, profitability? What does that pathway look like?

JAN SZILAGYI: So for us, obviously as a business, we're looking at engagement. And we're seeing quite a lot of not understandably at the moment, this is the moment of AI. And so I think there's immense amount of curiosity about how to use the product, and so on.

We're seeing a lot of-- the learning curve is extremely steep. We are looking primarily at how easy can we make it for market participants to make better decisions ultimately? And the way you then measure this afterwards is really by outcomes that they have for their portfolios. Are they able to improve relative to benchmarks that they set themselves or things like the S&P, and so on?

- How does your platform engage with a ChatGPT? Can we invest based on that?

JAN SZILAGYI: So it's a good question. I think on ChatGPT, ChatGPT is, of course, in the limelight because everybody is now learning how to use this technology. What we have taken is foundationally the same model, but we've trained it specifically and much more narrowly on financial analytics. Because the concern that you might have when you're using this technology in finance is, what if I get a wrong answer? There could be hallucinations. How can I be sure that what you're telling me is factually correct?

And so what you see with TOGGLE is that each time when some analysis is done and summarized for you, you have complete traceability or auditability down to where the data is that this conclusion came from so that you can quickly check. Just like if I tell you, hey, the price of something today is up by this much, it's nice to see the chart. Because you can say, OK, I can see that you're correct. It roughly looks like what you're telling me is true. And so we take the same approach with TOGGLE.

BRAD SMITH: So even as we think about what potential this may have in the future, first, I guess, can this allow somebody to outperform the markets right now? And then when we get to perhaps a baseline where the majority of portfolios or clients who are large banks and institutions are, including an AI approach within their own strategy, how does that not add more volatility than if we're all following the same recommendation?

JAN SZILAGYI: It's a good question. Obviously, on our side, the way we're thinking about this is the engine is going to be as good as the questions that you're asking of it. And so to an extent, it does elevate everybody's level of understanding of the market. But I don't think it necessarily changes dramatically the relative hierarchy. So there will be people who will be spending more time with it will continue to understand markets better, will use it differently than others.

I do think that just in the sense of financial literacy and a better understanding of how to avoid maybe some really bad mistakes, some really bad outcomes, that's where the technology can be very helpful.

- We will have to leave it there. Jan Szilagyi, thank you so much. TOGGLE AI CEO. Thanks for joining us.

JAN SZILAGYI: Thank you.