Creating an interoperable intelligence exchange for AI-ready networks

Commercial context

CSPs sit on immense volumes of intelligence, yet often cannot use it effectively. Data is locked in silos, shaped by legacy systems, fragmented operational logic, and strict privacy rules. These barriers suppress the value of network insights and prevent cross-CSP collaboration. They also make it difficult for CSPs to act on customer intent with speed or precision.

Commercial models remain dominated by static APIs and batch-driven campaign engines. These tools cannot match the pace of digital commerce, where users expect immediate and personalized interaction. As a result, CSPs rely on slow decision cycles and isolated engagement logic that cannot scale. Each organization builds its own AI stack, creating duplicated effort and inconsistent outcomes.

The industry needs a way to exchange intelligence without sharing data. It needs cross-CSP reasoning that respects sovereignty while enabling fast and secure decisioning. The challenge is to transform individual AI systems into a shared ecosystem that creates value for every participant.

The Agentic Intelligence Exchange (AIX) Catalyst is designed to solve this issue. Led by Telin, Huawei, Whale Cloud, and Valuoo, its goal is to demonstrate how distributed AI agents can connect insights, network APIs, and customer engagement tools in one continuous loop. The approach aims to develop new business value while supporting responsible AI use across the telecom sector.

AIX introduces a shared agentic AI hub that coordinates intelligence across CSPs and platforms. Instead of sharing raw data, each CSP deploys local agents that can sense context, run reasoning steps, and send decisions back into the hub. This creates a federated model that supports collaboration without compromising privacy.

The solution

The architectural foundation follows TM Forum’s Open Digital Architecture. It uses modular components that work across multiple networks and business domains. The system connects three layers of telecom capability: network access, AI reasoning, and customer engagement. Together, these layers form a closed decisioning loop.

ANeuAPIX, Telin’s CPaaS platform, forms the network-facing entry point of the decisioning pipeline. It provides the network interfaces and event streams that allow agents to interpret real-time context. This includes session activity, behavior signals, service usage, and other immediate triggers. These data points remain under each CSP's control. No raw data crosses the boundary.

The ecosystem is built on the high-velocity raw network data supplied by Indosat Ooredoo Hutchison (IOH). As the foundational operator, IOH unlocks the critical telco data assets including signaling logs, location telemetry, and session usage that serve as the ground truth for the entire exchange. This massive stream of sovereign data fuels the signal processing layer.

The pipeline then moves to the network edge, where Huawei’s real-time signal processing first verifies identity and context. Crucially, this layer provides the ‘Telco Verify’ capability, authenticating physical user identity and network trust before any data enters the AI pipeline. This ensures that the entire ecosystem operates on verified, spoof-proof signals.

These trusted signals feed into the AI engine and processor supplied by Whale Cloud. This acts as the cognitive core of the exchange. Instead of just aggregating data, it orchestrates AI agent reasoning that interprets intent, synthesizes a 360° customer context, and determines the next best action. The agents can request classification, prediction, or planning tasks across operators while maintaining strict privacy guarantees. Insights and decisions move, but raw user data does not.

The customer-facing layer connects through Valuoo’s Me2B engagement platform. Here, decisions turn into personalized actions across mobile, messaging, and digital channels. AIX automates activation, adapts messaging to context, and supports multi-CSP rewards and top-up flows. These capabilities enable behavior-driven campaigns based on intent rather than fixed segmentation.

The entire system is anchored in TM Forum assets that ensure both governance and interoperability. TMF915 provides the controls for managing AI models across their full lifecycle, supporting transparency and keeping every reasoning step auditable. TMF620 contributes by defining intelligence-based offerings within a standardized product catalog, while TMF622 automates how these offerings are provisioned across CSPs.

AIX maintains a consistent view of network and service context through TMF638, which synchronizes inventory data between participating CSPs. TMF680 adds the capability to provide personalized and context-aware recommendations, completing the engagement loop. Oversight comes from TMF951, which enforces transparent decision logging, bias mitigation, and broader AI governance. Together, these assets create a trusted, standards-based foundation for multi-CSP intelligence exchange.

Application and wider value

AIX provides new commercial value by allowing CSPs to monetize intelligence, not just connectivity. With this architecture, decision logic becomes a product, and insights become a service. Campaigns become cross-CSP assets that partners can activate in real time.

Telin’s NeuAPIX platform provides some compelling evidence. Here, intelligent engagement features already delivered a sixfold increase in click-through rate and threefold better conversion. By adding cross-CSP reasoning, AIX is expected to raise campaign efficiency by another 20–30%. Faster intent recognition reduces churn risk and shortens customer decision cycles.

Valuoo’s Me2B system benefits from the same shared intelligence. Personalized rewards and contextual top-ups become more accurate, raising conversion by 25–35%. Multi-CSP insights also speed up activation and cut friction in customer journeys.

Syahrier Firmansyah Wakid, Senior Product Developer at Telin and project lead, explains. "Operationally, AIX reduces onboarding time for AI services. Schema discovery powered by AI cuts integration work by as much as 90%. This moves new services from concept to launch within hours, not weeks. It also allows smaller CSPs to access advanced intelligence without building large data science teams. Overall, AIX has generated $3–5M revenue by turning telco data into real-time intelligence."

AIX proves that privacy-preserving collaboration is not only possible but commercially viable. CSPs can exchange intelligence while protecting customer data. They can support joint services without exposing datasets or losing control. The Catalyst also demonstrates how AI governance can scale across multiple organizations. Bias controls support trusted outcomes, and CSPs gain confidence that shared intelligence is logged, transparent, and meets compliance standards. These capabilities move the industry closer to AI-ready networks and support TM Forum’s goals for transparent, interoperable systems.

More broadly, AIX means people gain more relevant, respectful engagement. Authentication becomes faster and safer. Personalized experiences reflect consent and data boundaries. Services respond to intent without invasive data sharing. AIX shows that AI can be both powerful and human-centric, aligning commercial value with responsible practice in ways that strengthen trust and support growth in a rapidly evolving market.

The project won the Outstanding Catalyst – Business Gamechanger award at TM Forum’s Innovate Asia event in November 2025.