The AI Inference Opportunity for Telcos

AI is reshaping industries, organizations, and consumer experiences in unprecedented ways. Generative AI is evolving toward reasoning, agentic AI drives outcome-oriented workflows, and physical AI enables real-time decision-making at the edge—powering cameras, robots, drones, and autonomous systems to interact dynamically with their environments.

In 2017, mobile network operators (MNOs) had the perfect edge computing opportunity—hyperscalers needed to expand cloud regions, and telcos had the infrastructure to enable it. This should have been a massive revenue stream, yet edge computing remains an unrealized promise.

The issue was never technology; it was the business mindset. MNOs failed to craft a compelling value proposition, struggled to align with hyperscaler economics, and ultimately let the opportunity slip through their fingers.

Now, AI inference is here, and history is repeating itself. MNOs sit on millions of compute resources across central offices, base stations, and regional data centers—ideal for AI inference. Nvidia reports that 97% of telcos are investing in AI, yet only a few—like Singtel, Swisscom, Telenor, and SoftBank—are successfully monetizing it.

Meanwhile, AI infrastructure spending is exploding. Microsoft and Meta will invest $333 billion in AI data centers this year alone. Some mega-facilities house over 1 million GPUs, while MNOs own distributed infrastructure closer to users but fail to capitalize on it.

A Win-Win-Win for Enterprises and Carriers

Maximin offers MNOs turnkey AI-powered computer vision solutions that need inference resources. Our model provides:

  • Enterprises with low-latency AI deployment and data sovereignty without requiring fiber or Power over Ethernet (PoE).

  • MNOs with a new AI-driven monetization strategy, leveraging their distributed edge infrastructure for high-value AI inference workloads and gaining new revenue streams.

  • Maximin with the ability to scale its computer vision solutions, ensuring broad adoption and deeper enterprise integration.

To learn more, please get in touch with us.