With out context, AI guesses — creating unreliable outcomes. Ontologies intention to resolve the issue.
Throughout Cell World Congress Barcelona 2026, Totogi CEO Danielle Rios sat down with RCR Wi-fi Information’ Sean Kinney to resolve what’s stopping telcos from operating AI brokers, even after they have lots of of them deployed.
“It’s all about governance,” she mentioned. “They’re undecided they’re going to make the precise determination, so that they’re pumping the brakes.” The basis drawback, she defined, is context.
For instance, totally different telco techniques typically outline “subscriber” in another way. Is it a tool? Is it a line? An individual or an enterprise? AI, whereas highly effective at making selections, struggles in these ambiguous situations. When it lacks context, it guesses, resulting in the unreliable outcomes that operators fear about.
Rios mentioned that till telcos mannequin their whole operation — their enterprise processes, the choices they make, their guidelines — in a approach AI brokers can use throughout the enterprise, the issue will persist.
Totogi’s reply is an ontology — semantic framework that defines how knowledge, ideas, and relationships join inside a site. In contrast to easy classification instruments, they seize the relationships and context that enable techniques to interpret and act on knowledge, going past taxonomies or schemas that merely manage info.
The Totogi Ontology sits atop current BSS, OSS, and community techniques, unifying how what you are promoting is represented and enabling AI to grasp, cause, and act throughout your whole surroundings. By normalizing entities, processes, and actions from siloed purposes right into a single semantic mannequin, it creates a digital twin of your telco that permits AI to function reliably and execute duties with diminished hallucination threat. “It’s one illustration of your telco, one reality,” mentioned Rios. “While you create an AI utility, it runs by means of the ontology and all the way down to your legacy techniques. The ontology gained’t let it make a mistake.”
Actual-world deployments
Totogi says a number of operators are already making use of the method in real-world deployments. Almost 10 Tier-1 operators are presently utilizing its ontology answer.
One buyer, Zain Sudan, confronted a community challenge the place sure cells would go dormant with out triggering alarms. Earlier than implementing the Totogi Ontology, figuring out and resolving the problem usually took about 48 hours. “We introduced that all the way down to half-hour,” Rios mentioned. “And now we will add predictive AI on prime of it and cease it earlier than it even occurs.”
In distinction, Singapore-based StarHub is utilizing the Totogi Ontology inside its enterprise enterprise unit to speed up gross sales coaching and assist reside buyer conversations. As gross sales representatives converse with clients, the system transcribes the dialog, suggests responses, and recommends related merchandise. “It’s like having your CMO advocate the precise merchandise you need to promote, bundling the merchandise in the precise approach, with the very best margin,” Rios mentioned.
The method goals to protect the human component of enterprise gross sales whereas giving representatives real-time steering throughout conversations.
“As soon as telco begins to make use of that ontology and unlock use instances, we’ll actually see a hockey stick of adoption as a result of they’ll belief AI will make the precise determination,” Rios mentioned.