Ray Le Maistre, TelecomTV (00:08):
So we're back in Barcelona, but this time it's MWC26 and I'm here again with Danielle Rios, CEO at Totogi. DR, great to see you again. Here at MWC, we're here at the Totogi booth. It's in hall two and glad to see, of course, that the quality coffee is flowing as normal. Now, the conversation here at the show understandably is all about AI, Agentic AI. You spoke recently at the Agentic AI Summit and telcos are investing big in this area. But what are the biggest problems facing operators with Agentic AI?

Daniel Rios, Totogi (00:44):
Yeah. I recently had Rick Lievano from Microsoft on the Telco N20 podcast and shared this really interesting story, which is AT&T has developed over 400 agents. They stopped short of deploying them into production. And they were really worried about what was going to happen with them. Could they really trust that? Is the governance there for them? And so they didn't deploy them. And I think that's the real issue is how do you trust your AI? And the key is context. We're really focused on having great data, but no one's really talking about that enterprise-wide context that you need to build. And I think that's going to be the real conversation this year and in the coming year.

Ray Le Maistre, TelecomTV (01:19):
And why is context so hard for the telcos?

Daniel Rios, Totogi (01:22):
Yeah, I think everyone says garbage in, garbage out with AI. It's about the data. And I'm going to say it's about garbage context. You have hundreds of systems within that telco, all with different definitions of subscriber. And when AI doesn't know how to resolve those contradictions, it guesses. And that's the problem. With an ontology, with that enterprise-wide context, you can guide that AI so that it makes the right decision every time. And so that's... Our biggest problem is the issues with all the legacy systems, all the contradicting entities and processes. And so that's going to be the key with AI.

Ray Le Maistre, TelecomTV (01:59):
Okay. So how can Totogi help the operators with this context problem?

Daniel Rios, Totogi (02:05):
Yeah. So we've built something we call the Totogi ontology, and it is that one model that represents your telco. It's the one truth. And so what's so great about it is you connect any legacy system to it. It runs over your entire legacy layer and now your AI will act through the ontology. And so it can't hallucinate. It knows exactly your processes. It gives it the context it needs. It gives it all its decisioning. And what's so awesome, it compounds. So you can correct it, you can add to it. All sorts of things is pretty amazing and it's really, really powerful and we're really excited to show to telcos.

Ray Le Maistre, TelecomTV (02:40):
Okay. So your vision is one of an enterprise-wide context through ontology. But not talking theory here, you actually have several tier one telcos using this in their production networks. Can you tell us about those customers?

Daniel Rios, Totogi (02:54):
What's so great about the Totogi ontology is it works, right? And when people see it, they do deploy it to production. So we're starting to work with tier one telcos. We're up to almost nearly 10, but two we've named publicly, Zain Sudan and StarHub. So with Zain Sudan, they're using it on the network side to detect dormant cells. These are cells that don't really alarm on a dashboard. You don't really know that there's a problem with it until a subscriber calls you and says, "Hey, there's something wrong." It was taking them 48 hours to detect this problem. We've got it down to 30 minutes and now we can add some machine learning to it and actually prevent it before it happens. And so they saw that value and they're like, "Oh my gosh, this is like the Lego building blocks that really can advance our AI strategy.

(03:35):
It's what we've been looking for that gives us that confidence to deploy to production." With StarHub, on the other hand, they're using it in their enterprise group, right? They're helping sales reps to train them so that they know what products to pitch. They built a live transcriptor that's listening to a customer conversation, and it literally is a job aid that tells them not only what to say and what to sell, but actually the products to recommend. And so this is a CMO's dream, right? So we sit there and say, "Hey, don't sell this product, sell the higher margin product, bundle it with this." Literally give them a perfect script. It's like putting your CMO in every seat, your best sales guy in every seat talking to customers. So pretty powerful stuff.

Ray Le Maistre, TelecomTV (04:15):
Okay. So those sound like really meaningful results. What else is Totogi up to here at MWC26?

Daniel Rios, Totogi (04:22):
Well, we're here in our stand, which is amazing. We're showing operators how that ontology works and it's real. It's not a demo. So we're connected to real systems, real data and showing the power of that modelling of the enterprise-wide context. And when operators see that and they understand the struggles that they've been having with agents that are just connected to data but are missing the context, the light bulb goes off, they really see how powerful it is and they're like, "How do we get started with Totogi?" So really excited to do that.

Ray Le Maistre, TelecomTV (04:51):
Danielle, great to see you as ever. Thanks again for joining us on TelecomTV.

Daniel Rios, Totogi (04:55):
Thanks so much, Ray.