Friday 13 November 2026 · Vol. 1 · No. 017 SearchSign inSubscribe
Best Effort

The weekday briefing on Australian telecommunications, data centres and AI infrastructure.

Carriers & networks

TPG Telecom plans customer-triggered AI agents, reports 40 per cent lower churn

TPG Telecom is developing AI agents that customers could start themselves to diagnose and fix service problems, with general manager Gaston De Arriba reporting around 40 per cent lower churn in cohorts where agentic AI is already used. No launch date was given.

TPG Telecom is developing AI agents that customers could trigger themselves to diagnose and fix problems with their services, potentially without calling the telco, according to iTnews. General manager of technology operations Gaston De Arriba outlined the plan at Cisco Live in Melbourne. He said cohorts where TPG uses agentic AI have seen churn fall by around 40 per cent.

The agents build on Splunk tools that TPG deployed in March to make its service operations more AI-driven, an approach often called AIOps. TPG's stated aims included real-time visibility across its networks and IT environment and faster detection and resolution of issues. CIO Monitor reported that the March disclosure named Splunk but not individual products or versions.

As of Thursday 12 November 2026, the agents work in the background and support contact-centre staff rather than dealing with customers directly. They collect network performance data, KPIs, incidents, changes and complaints, correlate them and identify each customer's problem in real time. The system recommends what staff should tell the customer and proposes fixes for TPG to implement in the network. De Arriba said one of the main pain points was that customers call and the call centre "doesn't have enough information to tell them where is the impact and what we are doing to resolve it".

De Arriba described the churn result as an early metric. CIO Monitor noted that he gave no starting churn rate or cohort size, and that the figure does not isolate AI from other service interventions. He also said proactive action had deflected many calls to the call centre, an effect TPG had not expected, though no call volumes or savings were disclosed.

The next step is letting customers start the agents from their own services. De Arriba said TPG's appetite was "to give the power to the customer to trigger the agents from their own services, and then the agent to take care of the service performance for these customers". Customers could see service performance on screen, receive advice on what to change on their side, or raise a trouble ticket without phoning. CIO Monitor reported that no launch date, eligibility rules, coverage or commercial terms were disclosed.

De Arriba said TPG is moving slowly and keeping agents within defined boundaries. He named customer experience and security as priorities, the latter particularly as customers begin interacting with agents directly. Cost is a third risk: he called the "tokenomics" of deeper agent deployment a point of risk because agents can quickly raise operating costs.