Japan’s top mobile operator, NTT Docomo, is rolling out an agentic artificial intelligence system on a commercial basis to support maintenance and operations across its mobile network throughout the country. This could be a major step in the way mobile infrastructure is managed in the increasingly complicated telecom environments.
What Is the Agentic AI System?
The AI platform in question employs agentic AI technology an advanced form of autonomous AI which, among other things, can independently analyze, reason, and make suggestions for actions to handle network maintenance activities that previously had to be done manually. The system became functional on Docomo’s mobile network as of 4 February 2026, and is currently working in service areas or domains that are supported by multiple vendors and equipment types.
The rollout is based on one of the largest network datasets in the world that includes traffic metrics and alarm logs from more than 1 million devices, such as base stations and core equipment. By analyzing real, time data from various inputs, the AI is capable of identifying unusual events, determining the main cause of problems, and recommending the measures for rectification in the case of complicated failures.
Also Read: Datadog and Sakana AI Partner to Accelerate Enterprise AI in Production
Why This Matters
Mobile networks in the present day particularly those using 4G and 5G technologies are a lot more complicated than before, which means that faults must be detected and resolved very quickly. Prior to the AI platform, if issues were not related to the known conditions, engineers had to rely on predefined procedures or conduct extensive manual analysis. The AI system is able to shorten this cycle dramatically:
Automates root, cause analysis: The AI has the ability to comb through large datasets to find the components and patterns causing the problems that are not explained by the traditional playbooks.
Minimizes downtime: If the complex failure tests confirm the proposed response time, then it will be reduced by over 50 percent, thus network resilience and reliability will be greatly improved.
Facilitates autonomous operations: Docomo is progressing towards an autonomous network model, which is the main feature of 6G networks in the future by shifting from rule, based automation to AI, powered decision support.
How the Technology Works
Docomo built the system using Amazon Bedrock AgentCore to provide secure governance and scalability for the AI agents. Its architecture combines:
Multiple AI agents operating in tandem
Graph-modeled network topology data to understand relationships among network nodes
Time-series, tabular, and graph databases optimized for agentic workloads
This enables real-time detection of network abnormalities and supports proactive recommendations for maintenance teams—moving beyond simple alerts to intelligent troubleshooting assistance.
Strategic Impact
For Docomo, this initiative is one piece of a larger effort to leverage generative and agentic AI for the automation of the main operational processes. Through the integration of such systems in routine network operations:
The firm bolsters the dependability of mobile communications, which are indispensable for both consumer and commercial services, i. e. , among others, emergency response.
It lowers the company’s operational costs and raises productivity, particularly as the network becomes more complex with additional services and more connected devices.
It makes an AI, driven telecom operation, at which Docomo would be the leader, a pivotal technology for the future development of 6G.
Broader Industry Context
Docomo’s rollout is indicative of a bigger trend in the telecom industry whereby operators of telco networks turn to agentic and autonomous AI tools to run vast, multi, vendor networks more efficiently. Such solutions are progressively considered vital in cutting down on human intervention, improving the level of services, and allowing instant reaction to problems with the infrastructure, particularly, when the volume of data and the requirements of the services are constantly increasing.


