Intelligent telco operations through AI-driven, user-centric network insights

Commercial context

Customer loyalty increasingly hinges on network experience. According to McKinsey, around 20% of users base their decision to stay or leave a network provider on perceived performance. And as more services shift to mobile-first, the gap between network performance and actual user satisfaction grows.

While many CSPs invest in network upgrades, few can link those investments directly to what customers experience in real time. Capital planning often depends on theoretical coverage models or network KPIs, not the lived reality of users at the network edge. Meanwhile, generative AI and new content patterns are reshaping traffic profiles, especially with rising uplink demand and increased video generation from mobile devices. But without visibility into end-user performance, CSPs can only react after customers complain — or worse, after they leave.

The 'AI-powered end-to-end solution for customer experience' Catalyst has been established to overcome this. It builds a real-time view of customer experience across the network, designed not around systems, but around users. By combining AI-driven analysis with token (or blockchain)-based user incentive design, crowd-sourced insights, it gives CSPs a practical, scalable way to detect issues early, resolve them faster, and guide smarter investments.

The solution

The solution focuses on a simple idea: pay attention to the status of the user’s device. Through lightweight, anonymous data collection, the system gathers round-the-clock connectivity metrics — including coverage, latency, and capacity. It doesn’t rely on complaints or static dashboards. It observes real experience, as it happens.

To ensure privacy and transparency, the Catalyst uses a Decentralized Physical Infrastructure (DePIN) model. Users who contribute their network performance data are rewarded with tokens and personalized insights. This approach builds trust and participation without compromising privacy or creating new attack surfaces.

The system is efficient by design. Its footprint on the device is minimal, requiring little bandwidth and generating negligible network load. That makes the data collection process highly sustainable and affordable — a key concern for CSPs under pressure to reduce OPEX and carbon impact.

Once the system collects user-level data, it correlates it with infrastructure-based data from access, transport, and core network layers. Machine learning models combine both views to detect emerging issues, identify patterns, and surface high-impact bottlenecks. Instead of reacting to complaints, CSPs can pinpoint root causes, prioritize by customer impact, and address problems before they escalate.

Importantly, the system does more than generate alerts. It makes actionable recommendations based on real-world evidence and geotagged user data. For example, it might highlight specific device types struggling in certain bands or surface locations, including different building floors. These insights help CSPs fine-tune service delivery, adjust configurations, or redirect investment — always with customer experience in mind.

Behind the scenes, the Catalyst draws on TM Forum’s frameworks to ensure interoperability and operational readiness. Assets include the AI Use Case Guidebook (GB1002), the Data Governance Guidebook (GB1023), the Blockchain Use Cases for CSPs (TR279), and the Zero-Touch Operations toolkit. These assets help standardize implementation, support cross-system integration, and ensure CSPs can deploy the solution without needing to overhaul their architecture.

Wider application and value

At a business level, the impact is clear. By proactively resolving network issues tied to actual user experience, CSPs can reduce churn by an estimated 10–15%. This doesn’t just protect revenue — it boosts brand equity, raises NPS scores, and helps reduce the volume of inbound support calls.

Operationally, the system offers significant OPEX savings. Energy-efficient data collection and AI-optimized resource planning can reduce costs by 15–20%. Because field visits and manual diagnostics are minimized, CSPs also see benefits from fleet efficiency, road safety, and a reduced carbon footprint.

As Mr Okagawa, Senior VP General Manager, R&D Strategy Department at NTT Group, put it, “NTT DOCOMO will detect where improvements are needed at an early stage and implement appropriate measures quickly in order to achieve the No. 1 position in Japan in terms of mobile network experience assessment.”

The platform has long-term strategic value too. By tying customer data to network planning, it shifts CSPs toward more intelligent capex allocation. Investments are guided by impact, not assumption. That makes the network more responsive, and more aligned with how people actually use it.

At a societal level, the benefits ripple outward. Reduced downtime means more reliable access to services. Smarter resource use contributes to climate goals. And the token-incentive model gives users a role in shaping the quality of their own connectivity, flipping the relationship from reactive to collaborative.

With this Catalyst, CSPs can move away from assumptions about what the customer wants, to knowledge instead. Instead of tracking performance through systems, it watches how services behave in people’s hands. This user-first lens is the real innovation — and the foundation for smarter, more intelligent telecom networks.