Moonshot Catalyst: a giant leap in billing support

What do customers do when they don’t understand their bill or when they are billed far more
than they’d expected? They’re going to call their operator and ask for help.

Billing issues are one of the biggest causes of customer inquiries and dealing with these
inquiries takes a lot of resources and costs a lot of money. In fact, as Telecom Argentina’s Mariela Fiorenzo explains, making just one customer support call can render a customer unprofitable for the month (and maybe even the next one!).

Of course, customers will always need help, but the challenge is to meet these needs as efficiently and cost-effectively as possible, and that’s just what this year’s TM Forum AI Challenge Catalyst winner set out to address.

TM Forum Catalyst – AI-driven EBITDA Mastery: Revolutionizing Customer Journeys

This catalyst focused on answering customer inquiries more effectively while also minimizing the cost of dealing with them. The goal was to create smarter, automated customer care solutions that enable customer inquiries to be handled within digital channels (such as by smarter chatbots), as well as introducing smarter copilots within the call center to boost call center efficiency.

Two key technological enablers were employed: on-demand data exchange between ODA-compliant systems, which enables timely responses to AI / machine learning models; and function calling (TMF678) to locate and access data in order to answer customers’ queries.

To illustrate the effectiveness of the AI framework, the catalyst team concentrated on two use cases: sentiment-aligned AI billing inquiry management, which seeks to use the AI framework to improve customer service effectiveness and efficiency around bill inquiries; and field service management, which seeks to deliver a customer-centric approach to engineering issues while significantly boosting field engineering efficiency.

The catalyst was championed by TM Forum members Vodafone, Telecom Argentina and
ConvergeICT, and brought together leading vendors who each contributed their expertise such as Calvi’s partner Aria Systems who provided customer billing invoice data, AWS who supplied the AI management tools and foundational model, Nespon who contributed field service diagnosis management data and integrated AI flows, and Salesforce who delivered the engagement and orchestration layers.

To deliver on the catalyst’s goals, the team had to overcome the limitations that fragmented and
siloed data place on AI initiatives. And this is where Calvi came in.

The Calvi Platform easily integrates with diverse billing systems, creating a federated data model that allows service providers to fuel their AI initiatives without having to embark on extensive digital transformation programs to consolidate their data.

The critical role Calvi played in this catalyst was to process data from Aria Billing Cloud, and detect and offer invoice insights and anomaly detection via TMF678 integration for function calling. The importance of Calvi’s role was described by Brendan O’Brien, Chief Innovation Officer and Co-founder of Aria Systems. “Simply stated,” he said, “any organization attempting to leverage GenAI to bolster customer experience and support without integrating rich, fulsome, real-time billing data is bound to fail. Not just fall short. Straight up fail.”

Harnessing GenAI, predictive analytics and intent analysis

Sometimes customers spend more than they intend and when the bill arrives they are upset and want answers. This is a negative experience for both the customer and the service provider, who now has to deal with the fallout.

Although this situation is all-too-familiar, there is a better way of handling it.

The Calvi Platform can provide vital insight to service providers to enable them to pre-warn
customers with higher bills than normal and offer proactive care. These customers can be guided to a smart chatbot as their first port of call, which is able to understand the customer’s problem and find the relevant information to come up with a solution. Alternatively, the customer might ring their service provider and a smart IVR solution will triage their problem.

This resolution to the customer’s problem might include waiving or reducing the charge as a gesture
of goodwill, offering a more appropriate tariff to avoid the problem occurring in future, or providing advice to the customer on how to avoid such charges. For most customers the solution will be acceptable, and the chatbot helps them quickly and painlessly resolve their problem without the need to queue. From the service provider’s point of view this is a low-cost way of dealing with unhappy customers, while transforming their negative experience into a loyalty-generating interaction that potentially leads to additional revenue.

If none of the suggested options are acceptable to the customer, or if the customer still needs more help, they can be transferred to a customer care agent along with a full history of the interaction. This means they don’t need to repeat any information, because the human agent benefits from an AI-generated summary to get up-to-speed instantly as well as advice on the next-best action.

The aim is to provide a world-class service experience at low cost and at scale that leaves everyone happier thanks to the power of billing data, automation and generative AI.

Impressive results

The catalyst was a great success and was awarded the prize for best AI-enabled catalyst 2024. It not only demonstrated the efficacy of the approach taken but revealed substantial gains in terms of revenue increases of 3-9% and OPEX savings of 15-26%.

Telecom Argentina’s Fiorenzo says that by taking this approach: “Not only can we dramatically cut costs but we can also boost our revenue by offering customers exactly what they need, right when they need it.”

Check out the video on the Moonshot Catalyst: See here.