Login
Register
AI-led services demand instant provisioning, guaranteed outcomes and zero disruption. This award-winning Catalyst introduces a model-as-a-service control plane powered by agentic orchestration to help CSPs monetize AI traffic and deliver measurable experience outcomes.
AI traffic changes everything. Networks have to deliver outcomes, not connectivity
AI workloads are reshaping traffic patterns, service expectations and the economics of telecom networks. Enterprise customers increasingly define requirements in terms of business outcomes rather than infrastructure, expecting guaranteed performance for AI inference services and digital applications.
The Catalyst, Game-X: Game-changing autonomous network experience – Phase II, addresses this shift through the model-as-a-service control plane scenario described in TM Forum’s GB1085. It focuses on B2B and B2B2B scenarios involving two customer actors: enterprise IT teams that define available intents, quality of service and budgets; and agent developers and users who build and use agents on top of those network intents.
Traditional rule-based orchestration and cascaded intent models create tightly coupled chains between BSS, OSS and transport controllers. That rigidity slows provisioning, limits service agility and makes it harder for CSPs to monetize the growing market for AI traffic.
As Chetan Narang, Director of AI Product Strategy at Colt Technology Services, explains, “Game-X Phase II highlights a critical inflection point as AI-driven workloads reshape traffic patterns and redefine network expectations. For networks, this signals a shift from purely delivering connectivity to enabling measurable outcomes, where experience becomes the primary value metric.” He adds that value creation will increasingly depend on open, ecosystem-driven commercial models that support AI-native services and capture demand in an experience-led economy.
Game-X Phase II introduces a model-as-a-service control plane that allows enterprise customers to express business intent for AI inference traffic using natural language. Rather than cascading intent from business to service to resource layers, the architecture moves toward fully agentic orchestration across domains.
For enterprise IT users, the Catalyst demonstrates zero-wait and zero-touch planning and provisioning. For developers and users, it provides a control API and a zero-trouble assurance closed loop. AI agents dynamically interpret business intent, coordinate actions across service, network and resource domains, and support cross-domain planning without predefined ontologies or workflows.
Once services are deployed, assurance is maintained through a continuous observe-plan-act loop. The solution combines active probing, telemetry and an assurance graph to measure network and application health, while reactive and prescriptive AI agents support deep troubleshooting and resource optimization. The team also demonstrates how OpenTelemetry can improve transparency and observability, helping build trust in agentic systems.
The Catalyst draws on a wide set of TM Forum assets, including GB1085 Model-as-a-Service, TMF921 Intent Management API, IG1218 Autonomous Networks Business Requirements and Framework, IG1252 Autonomous Network Levels Evaluation Methodology, IG1253 Intent in Autonomous Networks, IG1256 Autonomous Network Effectiveness Indicators, GB1059 Autonomous Network Levels Evaluation Guidebook, GB900 ODA Blueprint and GB1027 Zero Touch Partnering Reference Architecture. The team is also developing a Distributed Reasoning Protocol and reasoning primitives for digital twin for decision intelligence.
By using the model-as-a-service control plane as a test service vertical, the Catalyst gives CSPs a practical way to explore AI traffic monetization. It also provides an industry-first test of the GB1085 recommendation in a live Catalyst setting.
The expected business impact is significant. The project targets two times faster time to market through agentic orchestration and a 30% improvement in operational efficiency through agentic closed-loop automation. The architecture is designed to be more flexible and resilient than traditional tightly coupled approaches, reducing single points of failure and improving service responsiveness.
For CSPs, this creates a path to new revenue streams based on AI-native services and outcome-based connectivity. For enterprise customers, it promises faster access to network capabilities aligned with business intent. For the wider industry, it advances a scalable blueprint for autonomous networks where service experience, not connectivity alone, becomes the unit of value.
Game-X: Game-changing autonomous network experience – Phase II was showcased at DTW Ignite 2026 as part of the Catalyst program, demonstrating how enterprise intent can be translated into real-time network actions using agentic orchestration.
The project was named winner of the Best Moonshot Catalyst – Autonomous Networks challenge at DTW Ignite 2026, recognizing its contribution to self-optimizing, self-healing and zero-touch network operations.
