AI agents are expected to soon make purchasing and scheduling decisions autonomously on behalf of humans.
But Michael Fanous, a UC Berkeley computer science graduate and former CareRev machine learning engineer, argues that these agents currently lack a key piece of the puzzle: the full context needed to truly understand the people they're being programmed to serve.
Fanas claims that machines currently struggle to discern whether a person's professional profile on LinkedIn, activity on Instagram, and public government records all belong to the same person.
To solve this, he teamed up with his father Emad Fanous, a veteran CTO, to found Nyne, a startup that aims to become the intelligence layer that helps agents understand humans across their digital footprint.
On Friday, Nyne announced it had raised $5.3 million in seed funding led by Wischoff Ventures and South Park Commons, with participation from several angel investors, including Applied Semantics co-founder and Google AdSense pioneer Gil Elbaz.
Given how effective Google's ad targeting is at identifying users, it might seem like Nyne is tackling a problem already solved by classic machine learning, but CEO Michael Fanous insists that's not the case. Google's “secret sauce” is its exclusive access to users' search history and cross-platform activity, a data advantage the tech giant never shares with outside agencies, he said.
For anyone else, “this is a strangely difficult problem to solve,” explained Nichole Wischoff, founder of Wischoff Ventures, a private VC fund that backed the deal.
Fanous told TechCrunch that Nyne is tackling the problem by deploying millions of agents across the internet to analyze the public's digital footprint and applying machine learning techniques to that data.
Nyne can triangulate information about individuals by looking at their activity on major social networks like Instagram, Facebook, and X, as well as apps like SoundCloud and Strava.
Then, as more consumer-facing businesses deploy AI agents, Nyne will help them gain a deeper real-world understanding of both existing and potential customers.
“We can give them all the information about who might be able to help them take the next right action,” Fanas said. “When you make all these connections, you get a pretty deep understanding of a person's interests, hobbies, how they think about very specific things,” he added.
According to Wischoff, the market for this data is huge and valuable for companies that use AI agents to reach customers.
“How do I find out you're pregnant and sell A, B, or C as soon as possible?” she said.
Previous generations of ad tech companies were able to collect some of this data, but Nyne intends to collect this more precisely for the agent world.
As for how the father-son duo works together, the CEO says he has an ideal partnership with his CTO and his father.
“I think being with your co-founders makes it easier to walk away when things don't go well,” Fanas said. “Even if I have to text him at 3 a.m. to finish the launch, I know he'll still love me the next day.”