The inputs for fintech-focused AI

What

Cumberland Labs is a Singapore-based incubator focused on real-world applications of blockchain technologies. Founded in 2022, Cumberland Labs is an offshoot of DRW, a leader in trading and global market making.

Why

According to Naveen Agnihotri, Cumberland Labs’s CEO and Co-Founder, both the blockchain and AI have the potential to fundamentally alter fintechs’ operations. The blockchain, Agnihotri argues, has “exponentially increased” the number of people who are using public keys and other encryption tools.

“We will be exploring fundamental use cases of what to do with this technology and this infrastructure that's already been created,” he said.

AI, meanwhile, has “really captured the public’s imagination” through technologies like ChatGPT. “It seems almost like magic,” Agnihotri said; AI has now drawn greater attention from business leaders, who wish to use the technology to their advantage.

How

Agnihotri, who has worked extensively in AI, fintech, and neuroscience, argues that business leaders looking to incorporate AI into their operations should look at the inputs that go into successful AI models.

“I don't think there is enough appreciation of how important data is to doing any kind of AI or data analytics in the future,” he said.

Generative AI works because of the trillions of data points it draws from; fintechs and others looking to adopt AI in their operations thus need to overhaul their operations surrounding data, and “need to capture every single thing that is going on.”

“You need to capture every single customer journey endpoint, every single internal data point, and you need to capture it in a consistent way across the different ways in which customers are interacting across the different worksheets that you have internally,” Agnihotri said. “You need to be able to put all of that in one place.”

While this may sound easy, more established companies have grown up with siloed data-collection and -storage processes. Some employees may even have exclusive control over the data within their purview, raising the risk of key data being lost if relevant employees leave suddenly. As a result, Agnihotri encourages fintechs interested in using AI to first hire tech leaders experienced in data ops; with those inputs, fintechs can then build out the right models for productive and profitable use cases.

“It's really all about data,” he concluded.