Network automation sits at the center of Nokiaโs latest collaboration with Google Cloud. The companies have expanded their relationship to bring Google's Gemini models into the Nokia Assurance Center, a network software platform designed for telecommunications providers. Through this initiative, Nokia has created six purpose-built AI agents intended to reduce operational complexity, speed issue resolution, and help operators advance toward highly automated network environments.
๐ Key Highlights
- Six AI agents will support network operations tasks
- Gemini models power the new agent framework
- Router and event triage agents are already operational
- SaaS launch is scheduled for September 2026
- Additional agents arrive through updates into 2027
The new framework addresses a growing challenge for telecom operators as network infrastructure produces larger and more complicated streams of operational data. Existing processes often depend on manual investigation of alerts and incidents, creating delays when identifying the root cause of service disruptions. By applying AI agents directly to raw network data, the platform aims to help operators quickly separate critical infrastructure problems from routine noise, enabling faster responses and reducing the impact of outages.
At the core of the platform is a collection of specialized agents with distinct responsibilities. The router agent coordinates activity across the ecosystem and manages interactions among the other agents. The event triage agent reviews active alarms and compares them with historical information to identify likely causes and evaluate impact. Supporting these functions are the KPI selector agent, which interprets network performance metrics, and the anomaly reasoner agent, which examines unusual behavior to determine whether an issue is legitimate or simply a false alarm.
Additional capabilities come from the action reasoner agent and dashboard agent. The action reasoner evaluates ongoing events and recommends remediation actions based on available automation catalogs, while the dashboard agent enables users to create visual reports and monitoring screens through natural-language requests. Nokia built the multi-agent framework using Google Cloudโs Agent Development Kit on the Gemini Enterprise Agent Platform, allowing the solution to operate on standard Google Cloud compute and storage resources while remaining compatible with existing customer environments.
Nokia has also introduced a governance approach it describes as โglass box autonomy.โ Rather than removing human oversight, the platform provides engineers with confidence-based recommendations before actions are approved and executed. For approved low-risk situations, the same architecture can support fully automated workflows. Nokia states that operators can benefit from faster troubleshooting, fewer false alarms, easier access to analytics, and lower software-related costs. The router and event triage agents are already functional, while the broader SaaS offering is expected to launch through Google Cloud Marketplace in September 2026, with additional capabilities rolling out across Nokiaโs portfolio through 2027.
๐ What This Means (Our Analysis)
This announcement highlights a practical shift in how network automation is being deployed within telecommunications environments. Rather than focusing on a single AI function, Nokia is introducing a coordinated system of specialized agents that can collectively interpret data, evaluate events, recommend actions, and support operational teams. The approach reflects an effort to manage increasing network complexity without relying exclusively on manual intervention.
The structure of the platform is equally notable because it combines automation with human oversight. By keeping engineers involved in approval processes while allowing low-risk activities to operate automatically, Nokia is positioning automation as an operational support layer rather than a replacement for decision-makers. The phased rollout strategy also suggests a focus on immediate deployment value while gradually extending capabilities across a broader set of network management applications.
๐ Our Take: As telecom networks become more data-intensive, AI-driven operational frameworks are poised to play a larger role in keeping infrastructure responsive, efficient, and easier to manage at scale.