Global telecommunications networks handle immense volumes of data, with millions of user connections generating over 3,800 terabytes per minute. This unprecedented flow, encompassing traffic data, performance metrics, and configurations, challenges traditional automation tools. At NVIDIA’s GTC conference, new developments were announced to revolutionize telco operations using large telco models (LTMs) and AI agents through the NVIDIA AI Enterprise platform.
NVIDIA’s partners, including SoftBank, Tech Mahindra, and Amdocs, are leveraging LTMs, specialized large language models trained specifically on network data. These models power AI agents that automate decision-making, enhance operational efficiency, and boost network performance. LTMs are optimized for real-time network event interpretation and continuous learning to adapt to new challenges, aiming to reduce downtime and improve security.
Companies are deploying these innovations to address complex network issues. SoftBank’s LTM automatically adjusts network configurations during events with massive traffic. Tech Mahindra utilizes NVIDIA’s tools to offer a comprehensive network insight platform, while Amdocs, BubbleRAN, and ServiceNow develop AI agents to enhance 5G deployment and network incident resolution. These advancements seek to accelerate AI adoption across telecommunications, delivering cost savings, increased productivity, and improved customer experiences.
