By Przemek Tomczak, Director of Industry, Partnerships & Alliances at Ocient
Mobile World Congress in Barcelona is always intense—but this year felt particularly packed.
Between walking from Hall 2 to Hall 7, over to Hall 4, then across to Hall 8.1 and back again, it sometimes felt like a fitness challenge as much as an industry conference. Over four days, I had more than 60 conversations with operators, vendors, startups, integrators, and security teams from across the telecom ecosystem.
The energy across the event was strong, but the tone of the conversations this year was noticeably pragmatic. Many discussions focused less on hype and more on the real operational challenges operators are trying to solve today.
Several themes came up repeatedly.
Cost Pressure Is Now Front and Center
In nearly every conversation, cost came up early.
Cloud spending, duplicated analytics platforms, and growing infrastructure complexity are under intense scrutiny. Many telecom teams are being asked to modernize their data platforms and analytics capabilities—while keeping costs under tight control.
This has made cost optimization a foundational requirement rather than a secondary consideration. For many organizations, the starting point of any technology discussion now revolves around a simple question: How can we scale analytics while keeping economics sustainable?
AI Is Accelerating How Applications Are Built
Another clear shift this year was how teams are using AI in day-to-day development.
Across operators and vendors, teams described using AI tools to generate queries, connect datasets, and build internal tools much faster than before. AI-assisted development is clearly accelerating how analytics workflows and applications are created.
But as development velocity increases, many organizations are encountering a new realization.
Data Foundations Are Becoming the Real Bottleneck
Almost every conversation eventually arrived at the same underlying question:
Do we actually have the right data foundation to support this?
The challenge isn’t just the sheer volume of telecom data. It’s also about whether the data is usable, governed and trustworthy at scale. Organizations are increasingly focused on issues such as:
- Data quality and trust
- Governance and traceability
- Domain context for telecom networks
- The ability to efficiently analyze extremely large datasets
Several teams described new investments in domain ontologies and structured knowledge layers designed to give AI systems the context needed to reason about telecom infrastructure and network behavior.
Without that foundation, even the most advanced AI systems struggle to deliver meaningful results.
Full-Fidelity Data Is Becoming More Important
Another recurring theme was a growing shift away from sampled or heavily aggregated data.
For certain use cases—particularly network security, operations, and automation—operators increasingly want access to high-resolution telemetry and longer historical windows. Understanding what actually happened in a network often requires detailed, full-fidelity data rather than approximations.
This shift is driving new architectural discussions across the industry. Organizations are evaluating how to store, process, and analyze far larger volumes of network data while maintaining performance and controlling cost.
The Big Takeaway
After more than 60 conversations throughout the week, one conclusion stood out clearly.
AI will absolutely accelerate innovation across telecom.
But the organizations that move fastest won’t simply be the ones adopting AI tools. They will be the ones that get their data foundations and economics right—building platforms that provide trusted data, strong governance, and the ability to analyze massive datasets without exploding infrastructure costs.
Those fundamentals will ultimately determine how far—and how fast—AI can take the telecom industry.