en.Wedoany.com Reported - On June 22, Google Cloud and Nokia announced an expanded partnership to integrate the Google Gemini model into Nokia's network software suite, Nokia Assurance Center. The two companies will develop six specialized AI agents for telecom network operations scenarios, helping operators reduce operational costs, accelerate network fault localization, and drive network operations toward higher levels of automation.
Nokia Assurance Center is Nokia's automated operations platform for telecom network and service assurance, primarily used to support operators in monitoring and analyzing complex network status, alarms, performance indicators, and service quality. With the coexistence of 5G, fixed networks, cloud-native core networks, and multi-vendor equipment environments, operators face continuously increasing data volumes and alarm counts, making traditional manual troubleshooting increasingly difficult to match network complexity.
The core of this partnership is introducing Gemini's multimodal reasoning capabilities into the network assurance process. Nokia will launch six specialized AI agents covering tasks such as intent understanding, alarm triage, performance indicator interpretation, anomaly cause analysis, repair action recommendation, and visualization dashboard generation. These agents can independently handle single operations issues or collaboratively analyze complex network events.
Among the six agents, the routing agent serves as the central orchestrator, understanding user intent and coordinating other agents while ensuring operations comply with operational boundaries; the event triage agent analyzes current alarms and compares them with historical patterns to determine root causes and impact scope; the KPI selection agent interprets complex network indicators, definitions, and measurement units, helping the system understand network performance changes.
The anomaly reasoning agent determines whether network behavior deviations are genuine faults or false alarms; the action reasoning agent matches current events with the automation catalog and provides repair recommendations; the dashboard agent supports operations personnel in quickly generating visual analysis pages and tracking screens through natural language. For operators, this multi-agent architecture can connect information originally scattered across alarm systems, knowledge bases, work orders, and engineer experience.
Google Cloud and Nokia emphasize that this solution does not completely eliminate human involvement. The system adopts a "glass-box autonomy" approach, where the action reasoning agent first acts as a recommendation layer, proposing repair solutions to engineers based on confidence levels, with human final confirmation at key control points. For low-risk, policy-compliant scenarios, the same architecture can support closed-loop automated execution while retaining execution records.
In terms of deployment, Nokia uses the Google Cloud Agent Development Kit and Gemini Enterprise Agent Platform to develop relevant capabilities. The multi-agent framework can run on standard Google Cloud compute and storage resources and can be deployed using tools such as Kubernetes and Google Cloud Storage to adapt to operators' existing cloud environments. This deployment approach helps reduce the cost for operators to introduce customized software overlay layers.
According to official Google Cloud information, the routing agent and event triage agent are already fully functional. The platform is planned to be officially launched as a SaaS model on Google Cloud Marketplace in September 2026, when operators can deploy the first batch of certified agents and use them with Nokia Assurance Center. The remaining more complex agents will be delivered through rolling software updates, extending to Nokia's Unified Inventory, Data Suite, and Orchestration network software portfolios, from the end of 2026 through 2027.
Telecom network operations are shifting from passive monitoring to proactive analysis. In the past, operators had to manually filter large numbers of alarms, cross-reference historical faults, invoke scripts, and gradually locate issues; with the introduction of AI agents, the system can more quickly distinguish between critical infrastructure faults and background noise, and present possible root causes, impact scope, and repair paths in a centralized manner. A Google Cloud press release states that these agents can reduce network problem processing time by 50% to 80%, with complex issues potentially shortened from hours to minutes.
However, telecom networks are high-reliability infrastructure, and automated operations still require strict boundaries. Whether AI agent recommendations are accurate, explainable, compliant with operator policies, and capable of handling multi-vendor network environments will affect their actual deployment effectiveness. Particularly in core networks, transport networks, and critical business scenarios, closed-loop automation requires thorough validation and cannot simply replace the operations responsibility chain with model judgments.
Key areas for subsequent observation will focus on the September launch progress on Google Cloud Marketplace, feedback from initial operator deployments, the actual troubleshooting effectiveness of the six agents, and whether Nokia can extend this capability to a broader range of network automation products. If the partnership progresses smoothly, Gemini's integration into Nokia Assurance Center could become an important case study in the transition from AIOps to autonomous network operations.
This article is compiled by Wedoany. All AI citations must indicate the source as "Wedoany". If there is any infringement or other issues, please notify us promptly, and we will modify or delete it accordingly. Email: news@wedoany.com
