Totogi Ontology goes live in seven telecoms deployments

Totogi said its Ontology has gone live in seven new telecoms deployments across five continents, with operators using it in network operations, sales, customer onboarding and business support systems.

The deployments were introduced over the past 12 months and include projects in Africa, Latin America, Asia-Pacific, Europe, Central America and Southeast Asia. According to Totogi, the work follows a common pattern: an operator starts with one use case, measures the outcome in production and then expands the scope.

Among the examples cited, an African operator reduced the time needed to resolve dormant cells from days to 30 minutes. Totogi said the system linked network, customer and revenue data and identified outages that existing monitoring tools had missed, including towers that had stopped carrying traffic while still appearing healthy in monitoring systems.

That operator has since expanded the deployment to a dozen additional use cases, Totogi said. It also said a sub-Saharan African operator cut weekly alarm noise by 97% and reduced mean time to repair by 75% after using the system to narrow 1.2 million weekly alarms to a smaller list of root causes.

In Latin America, a mobile operator is using the system to guide network capital expenditure decisions on a site-by-site basis. Totogi said the operator ranks investment choices by revenue and margin impact rather than capacity alone, using information drawn from network, billing and customer systems.

Elsewhere, a leading Asia-Pacific operator used the system to connect call capture, customer relationship management and configure-price-quote software. Totogi said the deployment was completed in four weeks and allows enterprise sales conversations to move into quotes and orders without manual re-entry by sales staff.

A Central American operator is using the system to support about 3,000 customer welcome calls each month, according to Totogi. It said an AI agent uses customer and account context to conduct the calls and write the results back into the operator's systems, replacing work previously handled by human agents in calls lasting five to eight minutes.

BSS projects

Totogi also described two deployments linked to business support systems. A tier-one European services company went live on Totogi's cloud-native BSS in eight weeks, and Totogi said the system cut operating costs by up to 50% and automated up to 80% of configuration and testing.

In Southeast Asia, a tier-one operator has adopted the Ontology as the basis for AI across its BSS estate under a multi-year agreement, Totogi said. The first function is already live in production and has reduced order-failure resolution times from eight to 24 hours to minutes.

Two of the seven deployments have already expanded beyond their initial use case and moved into multi-year engagements, Totogi said. It added that this reflects a shift among telecoms operators from limited AI trials towards broader use in live operational environments.

Operating context

Totogi said the Ontology sits on top of existing telecoms systems rather than replacing them. The company argues that AI tools struggle to operate reliably across network, operations support, business support, customer relationship management, charging and care systems when each system represents the business differently.

Its approach is to provide what it describes as a single executable model of an operator, defining both the meaning of data and the valid actions AI systems can take. Totogi said this structure allows operators to deploy projects in weeks rather than undertake longer replacement programmes for underlying systems.

The product is positioned as a layer that connects with software from suppliers including Amdocs, Ericsson, Huawei and Salesforce. Totogi said this enables operators to introduce new AI-driven functions in production while keeping their existing operational software in place.

Danielle Rios, Chief Executive Officer of Totogi, said the focus is on measurable outcomes in live settings rather than isolated tests. "Anyone can make AI look good in isolation. Production is where telco gets hard," Rios said.

Rios said the company was seeing a broader pattern emerge across operators' estates. "We're seeing momentum across the entire estate of a telco. That's the Totogi model: start with a business case, prove measurable value in production, then expand," she said.