The rapid rise of generative AI has become a pressure test for every enterprise. Businesses need to quickly analyze its potential and then leverage it to their advantage.
But credit reporting company Experian had a tool other businesses didn’t — an in-house innovation lab established in 2010. The company was able to rely on an established research practice, which accelerated prototyping and assessment process, according to Kathleen Peters, chief innovation officer at Experian.
“From the [lab’s] very beginning, we were researching and building all types of advanced analytics,” Peters said during a CIO Dive live virtual event on March 12. “That included machine learning, neural networks, different types of graphing technology – even early forms of generative technology.”
The research efforts helped Experian push to be “on the forefront in terms of our usage, testing and exploration with the forms of large language models and generative AI,” Peters said.
Initially, Experian’s innovation lab was founded to harness the expertise of PhDs, data scientists, software architects and engineers, and to tie emerging technology with improved company outcomes.
“One of the things I think is unique … is it’s really focused on collaboration,” Peters said. Researchers at the lab connect with business units across Experian to ensure ideas translate into better experiences for customers, clients and consumers.
With generative AI, the mission is no different. Last year the company launched Experian Assistant, a platform that enables businesses to conversationally interact with the company’s troves of financial data, reducing the need for organizations to rely on their own data science teams for insights.
The company is also experimenting with internal applications of generative AI such as coding assistance and recommendations for development support.
“Part of our strategy is understanding the capabilities around AI, using those ourselves, getting some of the tips and tricks, and then packaging that in a way that we can make it available to our partners and to our customers as well,” Peters said.
The AI pipeline
Enterprise leaders continue to grapple with moving AI pilots past initial exploration stages. Most businesses also struggle to find business value in generative AI, according to Informatica data published last month.
For Experian, the process begins broadly, Peters said. Employees can submit ideas for projects, and the business also looks closely at suggestions from universities and the startup ecosystem.
“In many cases, those lead to interesting partnerships, even investments sometimes by Experian, to bring that idea and that collaboration to life and what we might be able to do together,” Peters said.
Then, the company assesses the potential for AI projects to deliver outcomes in line with its business strategies, according to Peters. Regular business reviews allow Experian to narrow its scope so it can pursue the use cases that will deliver the strongest return on investment.
Agentic AI, the next frontier for enterprise AI applications, is on the company’s radar for the rest of the year.
“Properly coded agentic AI can perform tasks smarter and faster, in many cases, than humans,” Peters said. “It makes that data-driven decisioning possible. It adds a lot of automation, which not only leads to business productivity, but also convenience for whoever is the end user of what you’ve created. That’s very exciting.”
