“In the Loop” host Ejaaz Ahamadeen and Yahoo Finance’s Deirdre Bosa are joined by Jesse Zhang, Decagon co-founder and CEO, and Runway co-founder and CEO Cristóbal Valenzuela at Yahoo Finance’s 2026 Invest conference to discuss refocusing attention on tangible AI value and trying to deliver it
There’s anxiety which we can’t ignore. So, Jesse, maybe final question to you. What are you seeing in your actual business?
DekaCon works on sort of customer service and these agents they’re improving, they’ve made huge improvements this year. What are you seeing in terms of adoption? From non-tech companies too, by the way.
Yeah, I mean, I think connecting it to the jitters in the market. I think when this happens, there’s a huge refocus on, okay, what is the actual tangible value that’s being provided uh in just the AI industry overall.
That’s one of the benefits of being in the application layer is you are very close to the, you know, the actual end business value, especially if you’re working with, you know, folks outside of the tech industry, you know, the Fortune 500.
Um, and for us it’s it’s generally pretty direct, right? So we our agents are kind of sitting in the customer experience. They’re talking directly with customers and, you know, for the businesses we work with, the value that they’re looking at is like, hey, how much can we automate and that, you know, affects our bottom line, but also how much can we improve the MPS and the the customer experience, right?
So we work with one of the top uh three airlines in the US and they talked about in their earnings how, you know, they their AI experience was actually able to raise their NPS by 25 points.
Because it’s, you know, people don’t have to wait, it’s super customized, uh they can get whatever they need done.
Um and then in terms of like, you know, just closing on like exactly what we’re seeing changing in the experience. I think as the models have gotten better,
the big change we’ve noticed is that you you’re actually able to just improve the realism and improve the actual like, you know, how well, how close to a human or even better than a human can an AI agent perform.
And that’s both in, you know, how it speaks, you know, the quality of the voice, you know, how how intelligent it is at operating the internal systems.
And even just uh so we released a a new model a couple weeks ago on the voice side and we kind of built that ourselves to make it sound really like a a call center.
And just that alone in the AB test was able to raise the metrics by a significant amount. And so that’s that’s what we’re seeing the progress of the models really impact what we’re doing.
