- The rise of AI workloads demands optimized storage and networking to support centralized training and distributed inference
- Telcos are transforming their data infrastructure to real-time, AI-driven data pipelines to enable predictive network management and rapid decision-making
- New Fierce Network Research report covers how telcos are rebuilding their data infrastructure to optimize for AI
When I was 12 years old, my mind was blown by a story titled "Hawk Among the Sparrows," by Dean McLaughlin, about a nuclear fighter jet mysteriously transported back in time to World War I Europe.
Surprisingly, the jet from 50 years in the future proves to be nearly useless. Radar and guided missiles don't work against World War I planes made mostly of wood and cloth. And the air corps can barely find enough fuel to keep flying. Eventually, the futuristic machine proves to be a decisive weapon — but it's a struggle.
Today's telcos are hawks among the sparrows, armed with futuristic AI technology driving a fundamental shift in how networks are designed and managed, and promising new business models. But AI's potential is held back by networking built for a prior era.
Helping telcos navigate to modern networks has been an ongoing theme at Fierce Network Research; our reports analyze the transition to cloud infrastructure, the network as strategic asset, the unexpected evolution of SD-WAN and more.
In our latest Fierce Network Research report, we dig deep into essential network infrastructure that's often overlooked and taken for granted: Data and storage. Beyond bandwidth: Building AI-ready telco networks for the next generation of data demands
Traditional telco data architectures aren't up to the task of providing current data on network status — they're optimized for delivering data that's hours or days old. That's useful for future planning, but not for using AI to optimize networks in real-time. Likewise, network data is siloed by business functions, failing to deliver a complete picture of the whole network. Ånd some of that data is out-of-date and no longer accurate.
In our report, we talk with technology leaders at Telus, MetTel and Verizon about how they are re-architecting their networks for AI, building real-time high-velocity data pipelines that enable predictive analytics and rapid inferencing at the network edge.
Democratizing data access
Canadian operator Telus exemplifies the industry’s shift by modernizing its data estate over five years, consolidating and cleansing data to create a unified, self-service data pipeline that supports AI-driven operations and democratizes data access across the organization. This approach eliminates redundant data and enables faster, more accurate insights for network planning and optimization.
“We democratized the data. My job is not to be the middleman. The less that people come to me to get something, the more I know I’m doing a good job," Jaime Tatis, Telus Senior Vice President and Chief AI Officer, told Fierce Network Research in an interview.
Similarly, MetTel replaced siloed data with a streaming data architecture, allowing internal and external subscribers to access real-time network data, which is then archived for historical analysis. These modernized data pipelines are foundational to achieving business agility and operational efficiency in AI-driven telco environments.
“We put data on the stream, and whoever wants it, whether internally or externally, has to subscribe to that data stream. All of that ends up in a data lake for history, trending and reporting,” MetTel CTO Ed Fox said. “We continue to throw storage at that and increase it pretty regularly.”
And Verizon is seeing AI demand on the network shift from training, requiring centralized data, to inferencing, which requires a more distributed data model. “We’re already noticing pressure points,” Srini Kalapala, Verizon senior vice president, technology and product development, said. The carrier anticipates increased requirements for both compute and storage at the edge, to deliver low latency and reliability to increasingly demanding customers.
To learn more about how operators are rearchitecting their data and storage infrastructure to meet AI demands, download our free report: Beyond bandwidth: Building AI-ready telco networks for the next generation of data demands
And by the way, I read "Hawk Among the Sparrows" in a copy of Analog Magazine that my cousin loaned me. Here's the cover — isn't it great?