Login
Register
This Catalyst shows how CSPs can use agentic AI, crowdsourced experience data and multi-source network intelligence to reduce churn, optimize CAPEX and OPEX, and make customer experience the unifying metric across commercial and technology teams.
Making customer experience the engine of network investment
Network experience is one of the biggest drivers of customer loyalty, but many CSPs still make optimization and investment decisions using fragmented technical data, isolated drive tests and slow manual analysis. The result is a persistent gap between strong network KPIs and what customers actually experience, which can translate into stagnant NPS, rising churn and CAPEX that does not always improve customer outcomes.
The Catalyst project, AI-powered end-to-end solution for customer experience – Phase II, addresses that challenge by putting customer experience at the center of CSP decision-making. The project was named as a finalist in the 2026 Catalyst Awards, Attendees’ Choice category, and builds on the Phase I pilot to move toward commercial deployments, expanding from five to eight CSPs and demonstrating how AI-powered customer experience intelligence can connect network performance, investment planning and business growth.
Global telecom infrastructure investment reached USD428 billion in 2023 and is expected to surpass USD500 billion by 2026, while mobile data growth is forecast to continue rising strongly through 2030. Yet traditional network optimization approaches often rely on siloed OSS data, limited drive testing and reactive analysis. This makes it difficult for CSPs to understand where customer experience is deteriorating, why issues are occurring and which investments will create the greatest business impact.
The hardest data stream to capture continuously is often the customer’s real-world connectivity experience. Traditional crowdsourcing models have struggled with participant scale and consent management, while drive testing cannot deliver the statistical coverage needed to reflect everyday experience across locations, devices and usage patterns.
This Catalyst tackles that issue by using customer experience as the unifying metric across CTO, CMO, operations and finance teams. Its goal is to help CSPs move beyond traditional network availability measures and toward proactive, customer-centric outcomes that directly support retention, advocacy and more effective capital allocation.
The Phase II solution brings together five core capabilities: continuous collection of customer experience insights, multi-source data fusion, autonomous decision-making, multimodal access to intelligence and investment optimization. Together, these capabilities create an end-to-end platform that can detect customer-impacting network issues, identify likely root causes, recommend the highest-value interventions and validate whether investments improve experience after deployment.
At the data layer, the solution fuses real-time network performance from OSS systems, crowdsourced customer experience metrics and predictive traffic forecasts. It then applies agentic AI across the value chain to automate detection, analysis and recommendations. LLM-based and map-based interfaces make complex network and investment intelligence accessible to non-specialist teams, enabling stakeholders from customer care to finance to query the system and act on contextual, business-relevant insights.
The project also introduces a Network DePIN approach to support customer experience data collection, including mechanisms for rewarding participation and managing consent. This helps CSPs collect real customer connectivity experience 24/7 from end-user devices, with a level of statistical coverage that traditional field testing cannot match.
The Catalyst team has used a broad set of TM Forum assets to assess, structure and mature the solution, including the Artificial Intelligence User Stories and Use Cases, Data Governance Guidebook, Autonomous Networks Framework, Autonomous Networks Business Requirements and Framework, Autonomous Networks Technical Architecture, Autonomous Networks Reference Architecture, Autonomous Network Levels Evaluation Methodology, Autonomous Network Level Evaluation Tool, RAN Quality Optimization Questionnaire and Autonomous Networks L4 High Value Scenarios.
Using these assets and guidelines, the team has restructured the solution and assessed its maturity against autonomous networks requirements. The team achieved a significant improvement in autonomous network maturity from Phase I to Phase II and is working toward AN Level 4 through self-evaluation questionnaires that identify gaps for further development.
The business impact of the Catalyst is defined by the alignment of technical performance with commercial growth. By replacing labor-intensive drive testing with AI-driven virtual performance monitoring, the solution targets a 30–50% reduction in field testing costs. Through smarter investment planning and demand-based modelling, it aims to reduce CAPEX by 10–15% by directing spend toward interventions that measurably improve customer experience.
The Catalyst also targets a 15–20% reduction in network-related churn by identifying and resolving issues before they affect users, alongside 30–40% faster problem resolution through automated root-cause analysis. For customers, the expected result is a 5–10 point uplift in NPS as network latency, throughput and reliability translate into better experiences for high-bandwidth applications and everyday connectivity.
“Proactive and automatic customer issues resolution aiming at 10–15% improvement in customer retention and 15–20% savings in OPEX,” says Mr Okagawa, Senior VP General Manager, R&D Strategy Dept, NTT Group.
Mr Okagawa adds that the Catalyst is valuable because it “brings the customer back into the centre of the CSP decision making” by combining customer experience insight collection, multi-source data fusion, autonomous decisions, multimodal interfaces and investment optimization. The project team expects the solution to deliver meaningful improvements in OPEX efficiency, CAPEX optimization and customer retention.
As CSPs continue to invest heavily in 5G, fiber, cloud-native networks and autonomous operations, the ability to connect network performance with customer outcomes is becoming increasingly important. This Catalyst demonstrates a practical path for using AI to make that connection measurable, actionable and commercially relevant.
By shifting from reactive monitoring to proactive experience management, the solution shows how CSPs can make more accountable investment decisions, reduce operational waste and improve loyalty. It also demonstrates how agentic AI and autonomous networks can support business outcomes beyond technical automation, helping operators create a shared language between network teams, commercial teams and the customers they serve.
