Login
Register
AI-era services require guaranteed connectivity, edge computing, and dependable performance, but autonomy remains fragmented across business, service, and network domains. This Catalyst connects intent, Agentic AI, AIOps, knowledge graphs, and TM Forum standards to support faster service onboarding, closed-loop assurance, and new 5G AI service opportunities.
Connecting intent, intelligence, and operations for reliable AI-era services
Communications service providers are preparing to support a new generation of AI-ready services across enterprise and consumer markets. These services can combine guaranteed connectivity, edge computing, low latency, and service-level commitments, but delivering them reliably requires far closer coordination between commercial intent, service management, network operations, and customer experience.
Today, autonomy remains fragmented across these domains. Sales and operational processes are often disconnected, while service risks, incidents, and configuration changes still require significant manual coordination. This makes it difficult to move quickly from customer demand to a service that has been priced, technically validated, activated, and assured end to end.
The Catalyst project, Autonomy accelerated: Connected intelligence for reliable agentic operations - Phase IV, explores how holistic autonomy can connect these activities. It focuses on translating enterprise sales intent into service intent, validating feasibility across network and edge resources, checking subscriber device experience, and supporting dynamic pricing before a service is delivered.
The Catalyst combines intent-driven operations, Agentic AI, AIOps, knowledge graphs, subscriber device experience, and a common data foundation. Its aim is to connect business, service management, network operations, and customer experience information so that decisions can be made with a real-time, end-to-end view.
A central innovation is the integration of intent between the business and service layers. Enterprise demand can be captured as sales intent, translated into service intent, and checked against subscriber device experience, forecast network conditions, and available edge resources. The system can then validate whether the requested service is feasible, resolve the intent across operational domains, and support pricing based on the required service characteristics.
This creates a foundation for 5G AI slices that combine guaranteed connectivity, subscriber device experience, edge AI computing, and service-level commitments. The use cases include premium sports stadium experiences, emergency services, computer vision, remote healthcare, smart industry, and regulated AI workloads, where responsiveness and dependable performance are critical.
Agentic AI, AIOps, and knowledge graphs are used to predict demand, identify service risks, and recommend the safest next action. Closed-loop coordination spans orchestration, assurance, incidents, problems, and configuration changes, while AI-native operational experiences bring business, service, and network information together for more complete decision-making.
The project uses TM Forum assets to connect commercial processes with service qualification, orchestration and assurance. TMFC036 supports lead qualification, TMFC027 guided selling and pricing, and TMFC001 product information.
TMFC002 and TMFC003 support the path from business intent to service delivery. TMFC009 and TMF645 check qualification against forecast data and a digital twin, while TMF921 supports intent interpretation and reporting.
TMFC008, TMF638 and TMF639 provide inventory context. TMF633 supports service discovery, TMFC007 coordinates fulfillment, and TMFC050, TMFC060 and TMFC015 support recommendations, knowledge and usage information.
The architecture is also informed by TM Forum guidance on Agentic AI, Autonomous Networks, intent, generative AI and digital twins, creating a standards-aligned route from network operations to service management and monetization.
The project targets a scalable path toward Level 4+ autonomy and zero-touch operations. Its intended outcomes include faster service onboarding, lower manual effort, fewer service failures and SLA breaches, faster detection and restoration, improved resilience, and more responsive customer experiences.
The project team identifies three headline commercial targets: 80% growth potential in the business-to-business market, 90% faster time to market for new services, and 99% first-time-right service delivery. It also describes new revenue opportunities from 5G AI slices, edge AI services, and guaranteed connectivity.
Dr. Dang Dinh Trang, Director of Research & Development Center at VNPT Group says, “agentic AI can orchestrate the service lifecycle from enterprise demand to active 5G AI slices, while supporting autonomous recovery from physical faults. This creates a route to enterprise revenue and zero-touch reliability.”
For the wider telecommunications industry, the Catalyst offers a practical model for joining business intent, service design, network feasibility, and operational assurance. This can help CSPs move away from isolated automation and toward outcome-driven autonomy spanning the full service lifecycle.
By combining intent, Agentic AI, knowledge graphs, digital-twin validation, and TM Forum standards, the project aims to give operators a more reliable way to launch and assure AI-ready services. It also shows how autonomous operations can support commercial differentiation, allowing CSPs to offer guaranteed service characteristics while reducing manual coordination across business and operational domains.
Meet the team at Innovate Asia in Bangkok, 17-19 November 2026.
