AWS Financial Services Symposium 2026 | Ioana Boier, NVIDIA

Clips
More from AWS Financial Services Symposium 2026

Ioana Boier

Global Head of Capital Markets

NVIDIA

play_circle_outline Ioana Boier: Building AI Infrastructure for Capital Markets — Data Curation for Noisy, Heterogeneous Signals

play_circle_outline CPU-to-GPU Migration Enables Generative Market Twins and NASDAQ LOB Replicas for Fast, Intelligent Trade Execution

Info

Ioana Boier, NVIDIA

Ioana Boier

Global Head of Capital Markets NVIDIA

search

Rebecca Knight

>> Hello everyone and welcome back to the Cube's coverage of the AWS Financial Services Symposium here in New York City. I'm your host, Rebecca Knight. I would like to welcome to the show Ioana Boier. She is the global head of capital markets at NVIDIA. Welcome.

>> Thank you so much, Rebecca. It's great to be here.

Rebecca Knight

>> So, Ioana, you sit at a fascinating intersection, capital markets and AI infrastructure. Can you give our viewers a sense of your role at NVIDIA and what you're focused on right now?

>> Yeah. I lead our global strategy for capital markets, which means that I get to work with a lot of trading firm and particularly algorithmic trading firms and their broader ecosystems. So exchanges, data, analytics providers. And you're right, it's a really exciting time to be working in that space.

Rebecca Knight

>> So, Ioana, when you are talking to capital markets firms about AI, what are the conversations that keep coming up? What are the challenges that you're seeing time and time again?

>> Yeah. I would say there are three core areas where we hear questions, where we create consistent value and where we see impact. And these are data research and execution. And let me take you through them one by one. So on the data side, there is massive amounts of data. And we are hearing about AI handling internet scale data sets. In finance, we also have not only large scale data sets but also small scale data sets, which is a challenge sometimes to train models, but also the data can be very noisy, very non-stationary. So those are all challenges that we have to think about differently than AI at large and specialized data curation, data ingestion pipelines for that type of financial data. It's also about data heterogeneity. So there's not only market data, but nowadays, we're able to integrate data coming from news and from social media and from all sorts of other data sources supply chain that were before maybe used in a niche way in a siloed way and now it's possible to glean the interactions in the markets in a more comprehensive way with the help of AI. So that's the data side. The second point is the research and research is mainly about research velocity. It used to take a long time, maybe six months, maybe more to get to strategies that are ready to be deployed in production. And being able to shorten that cycle with accelerated compute is really important because the throughput of successful and robust strategies once they are deployed means that these firms are making more revenues and those revenues go back into compute and there's this virtual cycle of essentially continuous research that fuels these new strategies. And then finally, research is only valuable if you can deploy it in production. So execution is really important. We are hearing a lot of interest in not only being fast in execution, but being also very smart. So this combination of speed and intelligence and finding the right points along that trade off, that's very important for our customers.

Rebecca Knight

>> So let's bring this to life a little bit. Can you walk us through a customer engagement? Because the scale of what you're doing with customers is extraordinary.

>> That's true. And it usually starts with some bottleneck, whether it's something that they would like to accelerate, some GPU transition from CPU compute, but oftentimes, and what I find super interesting is when customers are trying to do something that couldn't be done before. So one of these examples is simulating markets more faithfully. It used to be that quantitative research was about making certain assumptions or looking at the historical information and try to think of what the future would look like based on the past. And that's never repeating itself. The past is not repeating itself. So one of these interesting examples is the work that we're doing with NASDAQ. NASDAQ is both an NVIDIA partner as well as an AWS partner and they had this brilliant idea of trying to create digital replicas of the market of the limit order books with the help of generative AI. They spoke about this at our GTC conference and just thinking about it, markets are really complex dynamical systems. Being able to faithfully model the structure of the markets is important both for exchanges internally, but also for their customers, for their subscribers, that the trading firms that are looking to test their strategies more thoroughly can take advantage of this kind of digital market twins.

Rebecca Knight

>> It's game changers.

>> Yes.

Rebecca Knight

>> So last question. Looking ahead across the capital markets landscape, where do you see AI as having the greatest impact in the next few years? And for the firms that are getting ahead now, what are they doing differently?

>> I think it's everywhere. It's along the trading pipeline. Of course, we talked about research and that's where we see the most compute and the most impact going in because you have to be generating those successful strategies and the faster you do it, the better. But data is important and execution is important. So it's really throughout. What I would say it's really exciting is to see high performance computing come together with artificial intelligence to deliver that value and that added alpha.

Rebecca Knight

>> Ioana Boier, thank you so much. A great, great conversation.

>> Thank you. A pleasure to be here.

Rebecca Knight

>> I'm Rebecca Knight. Stay tuned for more of the Cube's coverage of the AWS Financial Services Symposium.

Ioana Boier, NVIDIA

search

Rebecca Knight

>> Hello everyone and welcome back to the Cube's coverage of the AWS Financial Services Symposium here in New York City. I'm your host, Rebecca Knight. I would like to welcome to the show Ioana Boier. She is the global head of capital markets at NVIDIA. Welcome.

>> Thank you so much, Rebecca. It's great to be here.

Rebecca Knight

>> So, Ioana, you sit at a fascinating intersection, capital markets and AI infrastructure. Can you give our viewers a sense of your role at NVIDIA and what you're focused on right now?

>> Yeah. I lead our global strategy for capital markets, which means that I get to work with a lot of trading firm and particularly algorithmic trading firms and their broader ecosystems. So exchanges, data, analytics providers. And you're right, it's a really exciting time to be working in that space.

Rebecca Knight

>> So, Ioana, when you are talking to capital markets firms about AI, what are the conversations that keep coming up? What are the challenges that you're seeing time and time again?

>> Yeah. I would say there are three core areas where we hear questions, where we create consistent value and where we see impact. And these are data research and execution. And let me take you through them one by one. So on the data side, there is massive amounts of data. And we are hearing about AI handling internet scale data sets. In finance, we also have not only large scale data sets but also small scale data sets, which is a challenge sometimes to train models, but also the data can be very noisy, very non-stationary. So those are all challenges that we have to think about differently than AI at large and specialized data curation, data ingestion pipelines for that type of financial data. It's also about data heterogeneity. So there's not only market data, but nowadays, we're able to integrate data coming from news and from social media and from all sorts of other data sources supply chain that were before maybe used in a niche way in a siloed way and now it's possible to glean the interactions in the markets in a more comprehensive way with the help of AI. So that's the data side. The second point is the research and research is mainly about research velocity. It used to take a long time, maybe six months, maybe more to get to strategies that are ready to be deployed in production. And being able to shorten that cycle with accelerated compute is really important because the throughput of successful and robust strategies once they are deployed means that these firms are making more revenues and those revenues go back into compute and there's this virtual cycle of essentially continuous research that fuels these new strategies. And then finally, research is only valuable if you can deploy it in production. So execution is really important. We are hearing a lot of interest in not only being fast in execution, but being also very smart. So this combination of speed and intelligence and finding the right points along that trade off, that's very important for our customers.

Rebecca Knight

>> So let's bring this to life a little bit. Can you walk us through a customer engagement? Because the scale of what you're doing with customers is extraordinary.

>> That's true. And it usually starts with some bottleneck, whether it's something that they would like to accelerate, some GPU transition from CPU compute, but oftentimes, and what I find super interesting is when customers are trying to do something that couldn't be done before. So one of these examples is simulating markets more faithfully. It used to be that quantitative research was about making certain assumptions or looking at the historical information and try to think of what the future would look like based on the past. And that's never repeating itself. The past is not repeating itself. So one of these interesting examples is the work that we're doing with NASDAQ. NASDAQ is both an NVIDIA partner as well as an AWS partner and they had this brilliant idea of trying to create digital replicas of the market of the limit order books with the help of generative AI. They spoke about this at our GTC conference and just thinking about it, markets are really complex dynamical systems. Being able to faithfully model the structure of the markets is important both for exchanges internally, but also for their customers, for their subscribers, that the trading firms that are looking to test their strategies more thoroughly can take advantage of this kind of digital market twins.

Rebecca Knight

>> It's game changers.

>> Yes.

Rebecca Knight

>> So last question. Looking ahead across the capital markets landscape, where do you see AI as having the greatest impact in the next few years? And for the firms that are getting ahead now, what are they doing differently?

>> I think it's everywhere. It's along the trading pipeline. Of course, we talked about research and that's where we see the most compute and the most impact going in because you have to be generating those successful strategies and the faster you do it, the better. But data is important and execution is important. So it's really throughout. What I would say it's really exciting is to see high performance computing come together with artificial intelligence to deliver that value and that added alpha.

Rebecca Knight

>> Ioana Boier, thank you so much. A great, great conversation.

>> Thank you. A pleasure to be here.

Rebecca Knight

>> I'm Rebecca Knight. Stay tuned for more of the Cube's coverage of the AWS Financial Services Symposium.