PRISM-AI Phase II delivers a compelling evolution from fragmented, reactive operations to trusted, autonomous, closed-loop service assurance across multi-domain and multi-CSP environments. Storyline: Enterprises increasingly rely on AI services for mission-critical operations, yet service performance depends on complex interactions between Edge Cloud (TEN/SEN), IP, GPON, All Photonic Network(APN) domains. Today, these layers operate in silos, resulting in slow root-cause identification, high MTTR, and SLA breaches. PRISM-AI addresses this by introducing an agentic AI-driven framework capable of continuously monitoring service KPIs, correlating them with infrastructure telemetry, and autonomously resolving issues—even across CSP boundaries. NEC, Xacria and Qeema are providing their solutions to address these problems. Solution: The Catalyst implements a closed-loop, intent-driven architecture built on TM Forum Open APIs and ODA components. It combines: * Qeema AI Agents (Correlation, Decision, Intent) for multi-domain fault detection and root-cause analysis * XCLADRA + Qeema Agentic AI for real-time KPI evaluation, alarm correlation, and resolution planning * TMF APIs (TMF641, 642, 628, 657, 730, 921) enabling standardized service lifecycle, assurance, and automated remediation * ENNI-based multi-CSP integration includes KDDI, Telin, Mauritius Telecom, NTT and Fibercop to enable secure, minimal-information exchange across operator boundaries. The system autonomously determines whether service degradation originates from Edge resources, transport, or external CSP domains, and executes corrective actions such as rerouting traffic, initiating intent-based remediation, or inter-CSP coordination. Deliverables: * End-to-end Agentic AI-powered assurance platform prototype * Multi-CSP ENNI-integrated service assurance workflows * Demonstrated closed-loop automation (detect → correlate → decide → act) PRISM-AI demonstrates a scalable, standards-based blueprint for Autonomous Networks (Level 4), positioning CSPs to deliver reliable, differentiated Edge AI services. PRISM-AI Phase II has the potential to fundamentally reshape future telecom operations by introducing a new paradigm of autonomous, collaborative, and service-centric networks. First, the Catalyst accelerates the transition toward Level 4 Autonomous Networks, where decision-making and execution are no longer human-dependent but driven by agentic AI operating across business, service, and resource layers. By demonstrating real-time closed-loop automation—spanning detection, correlation, intent generation, and execution—it establishes a practical blueprint for fully self-managing networks. Second, it redefines how multi-operator ecosystems function. Today, cross-CSP service assurance is slow, manual, and constrained by data-sharing limitations. PRISM-AI introduces a federated AI collaboration model, enabling CSPs to exchange only the minimum required insights over standardized ENNI interfaces while maintaining autonomy and security. The Catalyst also demonstrates API-driven exposure of network topology, monitoring, and service assurance information across participating CSP domains, enabling real-time observability for AI-driven autonomous operations. This model could become the industry standard for inter-operator trust and automation, especially as networks become increasingly interconnected. Third, the integration of next-generation infrastructure such as All Photonic Networks (APN) into AI-driven assurance represents a major leap. By embedding optical-layer intelligence into automated decision loops, the solution enables dynamic, real-time path optimization across photonic and IP layers, paving the way for ultra-low latency, high-capacity services required by future applications such as AI and Industry 4.0. Fourth, FiberCop Edge Cloud Nodes enable a distributed, AI-native infrastructure that brings compute, storage, and networking closer to the edge for real-time workloads. Integrated with AI agent-driven assurance, they enable proactive diagnostics of service degradation, allowing early detection, isolation, and mitigation of performance issues across compute, transport, and service layers. This approach delivers added value to customers requiring reliable AI applications running on Edge nodes, ensuring higher resilience, lower latency, and consistent service quality in autonomous networks. Finally, PRISM-AI shifts the industry from network-centric to service-experience-centric operations. By correlating application performance, edge cloud metrics, and network telemetry, it enables CSPs to directly align operations with customer experience and business outcomes, not just infrastructure health. TMF API Resources Service Catalog TMF633 Resource Catalog TMF634 Service Quality TMF657 Service ordering TMF641 Resource ordering TMF652 Performance monitoring TMF628 Alarm exposure TMF642 Event subscription TMF641/TMF688 Intent Management TMF921 ODA Components TMF C006 Service Catalog, exposing certified TMF633 APIs TMF C007 Service Order Manager, exposing certified TMF641 APIs TMF C008 Service Inventory, exposing certified TMF638 APIs TMF C010 Resource Catalog, exposing certified TMF634 APIs TMF C011 Resource Order Manager, exposing certified TMF652 APIs TMF C012 Resource Inventory, exposing certified TMF639 APIs