.png)
Last week in Copenhagen, at TM Forum's DTW Ignite 2026, the Wholesale Broadband as a Service Catalyst Phase III was named a finalist for the Outstanding Catalyst Business Gamechanger Award at the TM Forum Open Innovation Catalyst Awards 2026.
It is recognition for the consortium, and for work that has been years in the making. But award or not, the problem this project is solving is worth understanding on its own terms.
Every broadband customer knows the experience. The connection drops, the router lights do something unhelpful, and you call your provider. So does everyone else. The wait begins, and nobody tells you anything useful. It is not a new problem. But in 2026, the answer is changing, and it goes well beyond putting a chatbot on a help page.
AI in telecom customer service is not new. What comes next is. Most large operators already use some form of automated support. Chatbots handle FAQs, route calls, check account status. That part is broadly in place across the industry.
The harder problem is structural. When a broadband fault is not on the customer's device and not on the service provider's systems, it sits somewhere in the underlying network infrastructure, operated by a separate company. In wholesale telecoms, that boundary between the service provider and the network operator is where resolution slows down. Tickets are passed manually, coordination happens over email and phone calls, and customers wait longest precisely where the fault is hardest to trace.
That is the problem worth solving. And that is what this phase of the Catalyst is built to address.
As part of the TM Forum Catalyst programme, a consortium of network and service providers has been working on a system that automates fault resolution across that boundary, end to end, standards-based, and built to work across different operators and vendor environments.
The project, Wholesale Broadband as a Service now in Phase III, focuses on how to implement agentic AI in wholesale telecoms today. And Alvatross is part of this consortium, alongside partners of network and service providers.
And how does this work? When you report a broadband problem, you contact your service provider (in other words, the company you have a contract with). But the physical network (the real cables, the fibre, the access equipment…) often belongs to a separate operator working in the background. Coordinating between the two has historically meant manual handoffs, slow escalations, and customers left waiting with no information.
The system built in this Catalyst automates that entire journey. So, now, when a customer opens a chat and reports a problem, three AI agents take over. The first keeps the customer informed throughout. The second works on the service provider side, running diagnostics, ruling out faults on the customer's own equipment, deciding whether the fault lies deeper in the network. If it does, the third agent operates on the network operator side: running independent diagnostics, identifying the root cause, and scheduling remediation.
The handoff between the two organisations (traditionally the slowest part) happens automatically, across open TM Forum APIs, without a human needing to manage it.
The result: fault resolution that used to take days now takes hours, or minutes. OPEX cost per trouble ticket reduced by over 30%. NxM vendor integrations replaced by a single reusable profile, with bespoke integration costs starting at a €500k saving.
Telecoms does not run on tidy, single-vendor infrastructure. Any real network involves dozens of systems, platforms and providers that all need to communicate.
This solution is built on TM Forum Open APIs and the ODA framework, which means it is not a custom build for one operator that cannot travel. It is a reusable blueprint, proven in a multi-vendor environment, that other operators can adopt without rebuilding from scratch. The architecture feeds directly back into the TM Forum Wholesale Broadband Standardisation member project, strengthening the standard itself.
That is what makes the efficiency gains credible and repeatable, not just demonstrable in a controlled setting.
Alvatross has been part of this Catalyst from the beginning. Our platform is built on the same TM Forum standards that make cross-domain coordination like this possible, one architecture designed to orchestrate across network types, operators and boundaries without requiring custom integration at every join.
The problems this project addresses are the problems we hear about in real conversations with real clients. Operators are not asking for more complexity. They are asking for smarter ways to manage what they already have; across systems and partners they did not choose in isolation.
You can find the full project details on the official TM Forum Catalyst page: Wholesale Broadband as a Service, The Future of Service Assurance, Phase III.
