• by Laurie Sullivan , Staff Writer @lauriesullivan, Yesterday

Gudea, a spinoff from a non-profit, has adapted strategies and tactics that it used to support intelligence communities overseas to disrupt online Al-Qaeda and ISIS networks, stopping information from spreading across the internet.

The startup maps nodes -- connection points for devices running across the internet -- to identify how information moves. The patented technology can identify about 10 million interactions.

Information is ingested across 47 platforms. The social platforms do not interact in silos. Information from YouTube moves into Facebook, X and then into Reddit, creating an entire ecosystem.

The technology then plots how everyone interacts with the platform and runs AI to identify patterns and make predictions.

Keith Presley, CEO of Gudea, calls it graph neural networking. “We built technology that allowed us to map how information moves online,” Presley said. “After using it for a few different campaigns, we realized the patterns repeated.”

By layering AI on top of this, Gudea found it could predict the movement of patterns based on what occurred in the past. That unlocked a slew of different things the company could offer clients.

Gudea works with Fortune 50 clients and PR firms in telecom and retail, but Presley declined to identify them. The focus today remains on reputation and risk management, but he believes there is a place for offerings related to predictive advertising and technology. 

With social listening, for example, the technology analyzes patterns and how people interact with information. Searches based on terms and keywords are not needed. The technology can find what Presley called "damaging words and phrases" without knowing exactly what companies or brands should look for. 

Presley approaches influencers differently than influencer strategies. The technology has learned to reach much further than what brands believe they should.

Take the topic of DEI, for example. “We will know where this topic will surface long before it gains any traction,” Presley said. “We map how the information will spread, so we can identify and tell the brand where it will surface based on layers of the audience. Something in 5-, 10-, or 20-layer audiences.”

Layers refer to predicting how wide and deep across the internet the information will spread. The technology can identify deeper audience layers, up to 70. To mitigate any negative impact,

Presley advises that the brand or company advertise to the outer audience layer, such as 70, so by the time the information gets there, the damage has disappeared.

There could be as many as 70 audience layers. How does Gudea define an audience layer? "We’re building out connection based on how a person interacts with the information over time," he said. "It's the next step how the person interacts with the information.”

The startup is putting together a road map for its products focused on advertising and marketing.

“Because we’re mapping how individuals interact with information online, it creates these patterns of interaction,” Presley said. “So anytime it creates coordinated behavior happens within the network, it creates beautiful patterns that are easily picked up.”