- Mavenir’s John Larson highlights misconceptions about AI in telecoms, stressing practical, problem-driven applications.
- The company advocates AI-driven automation without requiring extensive GPU investments or large data lakes.
What happened: Mavenir pushes practical AI for telecom automation
At MWC25, Mavenir stressed the need for a practical AI approach in telecommunications, challenging the belief that AI deployment requires large-scale GPU investments and extensive data lakes. John Larson, Senior Vice President, highlighted widespread misconceptions, noting that many associate AI primarily with Generative AI (Gen AI) and Large Language Models (LLMs).
Instead, Larson explained how Mavenir integrates AI into telecom networks using existing infrastructure. The company applies machine learning techniques like XGBoost for fraud detection and security monitoring, avoiding the heavy computational demands of LLMs.
He also detailed how AI-driven automation optimises network operations, reducing manual tasks in workload deployment, software updates, and performance management. By focusing on solving real-world challenges rather than adopting AI for its own sake, Mavenir aims to enhance operational efficiency.
With Gen AI applications increasing data traffic, some argue that more AI is needed for network management. However, Larson emphasised the importance of leveraging existing network data effectively before turning to complex AI models.
Mavenir’s strategy aligns with the industry’s shift towards cloud-native automation, integrating AI into Kubernetes-based control planes for self-regulating networks. This approach enhances efficiency and scalability while minimising costs.
Also Read: Telefónica Germany upgrades to Mavenir’s cloud-native IMS
Also Read: EOLO expands 5G FWA network with Mavenir
Why it’s important
The telecom industry is rapidly adopting AI and automation to manage complex 5G networks, yet misconceptions persist. Many assume advanced AI requires large-scale computing resources, but Mavenir advocates for efficient AI integration within existing infrastructure.
For operators, this offers a cost-effective solution. Instead of heavy investments in GPU clusters, AI techniques like XGBoost can address network security, fraud detection, and automation, boosting efficiency without major hardware upgrades.
As Gen AI applications drive higher data traffic, Larson warns against a technology-first approach, urging operators to define clear problem statements before deploying AI.
Mavenir’s focus on cloud-native automation aligns with industry trends, where Kubernetes-based AI frameworkssupport next-generation networks, shifting towards practical AI deployment to solve real-world telecom challenges.
