Nasdaq CEO: GenAI ‘a whole new opportunity’

Nasdaq is using generative AI in several areas across its business, and aims to put AI tools in the hands of all of Nasdaq employees, said CEO Adena Friedman at the exchange’s Investor Day, held Tuesday in New York.

In addition to Friedman’s verbal embrace of genAI, other Nasdaq execs explained the benefits that cloud, open-source technology, and large language models have brought to their respective businesses.

The exchange has been working with AI for several years, but Friedman said the firm’s innovations with emerging technologies rely on a data-driven foundation and that Nasdaq’s advancements in AI have been underpinned by the company’s prior commitments to data governance and migrating data to the cloud. 

All of Nasdaq’s data now sits within a cloud framework with modern data management, Freidman said. “Generative AI is a whole new opportunity for us,” she said.

We’re leaning into generative AI. We’re using large language models to reduce time spent on research and documentation, and believe that will have an attractive ROI

Hazel Dalton, Nasdaq

Brenda Hoffman, CTO of the financial technology and market services division, explained that in 2008, the exchange housed all data from each of its covered markets and platforms on premises, and it had become increasingly difficult to manage. This resulted in the creation of Nasdaq’s data infrastructure and data lake. 

“Everything that goes into that data lake has to be standardized and normalized, so if you roll forward to today, we are pumping over 200 billion messages a day into that data lake,” Hoffman said. “That’s four times what it was in 2020, and we’ve been able to scale with that. With the standardization of that data, we’ve been able to ensure we build common modern APIs to access that data that’s allowed us to build products quickly.”

Over the last year, generative AI has been lauded by industry professionals for its potential to cut the time people spend on menial tasks, allowing them to pursue more important ones. Hoffman said Nasdaq’s generative AI implementation has already resulted in time savings. 

For example, by using Code Companion—ChatGPT’s coding partner that helps write, improve, and test code—Nasdaq has been able to run huge regression betas and use AI to understand the results. And Hoffman said Nasdaq is pushing all of its product documentation specifications into large language models (LLMs).

Going to the source 

Brad Peterson, CTO and chief information officer at Nasdaq, said the exchange has made significant progress on the AI front, and that the technology is attractive from a revenue standpoint. He traces Nasdaq’s interest in it to the 2000s, where early, open-source language models grew popular with the expansion of the web.

Those primitive models became the foundation for OpenAI’s ChatGPT LLM-based chatbot, and its competitor Anthropic with its “safer” LLM, Claude—AI products that were unimaginable in the early 2000s—as well as countless AI-based platforms, tools, products, and services in finance and beyond.

“Our view is that as open source continues to follow those advancements, we see the open-source models continuing to offer a very attractive alternative way of putting anything high-volume into a product,” Peterson said. “We think open source is the way to go.”

Verafin, a provider of anti-financial crime solutions that Nasdaq acquired in 2021, is also making progress with the use of internally developed LLMs, said Hazel Dalton, CTO of financial crime management technology at Nasdaq.

“The data that we have available to us allows our machine-learning models to shine because trading data feeds machine-learning models, and you get better results,” she said. “We’re leaning into generative AI. We’re using large language models to reduce time spent on research and documentation, and believe that will have an attractive ROI.”

Cumulo-nimble

Peterson pointed to the close relationships Nasdaq has with Big Tech firms as the reason for the exchange’s technological success. 

“We have a strategic advantage in that we have unique access and relationships with the best technology companies in the world,” Peterson said. “We get to work with AWS, Intel, AMD, Microsoft, and Nvidia, going deep with them on finding strategic projects that are meaningful to both of us, and there’s very little competitive overlap.”

Nasdaq’s high-profile partnership with cloud provider AWS allowed it to begin migrating its markets onto the cloud, and last year the exchange announced it had moved a second matching engine onto AWS. For Nasdaq’s Hoffman, the considerable progress the exchange has made with the migration effort was once a pipe dream. 

“We are the first global exchange operator to take a major regulated market to cloud technology, and we did that back in 2022,” she said. “Everybody back then said it couldn’t be done. We assembled a team of engineers, businesspeople, and lawyers, and with AWS, set out on a journey. In just under 11 months, we were launching that market on cloud infrastructure. We created a blueprint.”