P.I. Works software intelligence sits above the hardware to automate network management and improve performance.
At MWC 2026 Barcelona, the industry’s premier event, Sushil Rawat, Director of RAN strategy at TELUS and Djakhongir Siradjev, CTO of P.I. Works, discussed how they are moving beyond “black box” infrastructure to redefine network intelligence in an interview with RCR Tech’s Principal Analyst Sean Kinney.
Beyond the hype of future technologies, P.I. Works is delivering proven, AI-driven automation for the complex 5G Standalone networks of today. Through its EVO platform and EXA applications, P.I. Works provides the “bare-metal” visibility and cross-layer intelligence required to manage sophisticated Open RAN deployments. This collaboration with TELUS proves that independent, software-defined management is the key to unlocking new revenue streams right now.
TELUS’s open RAN deployment
TELUS has moved beyond pilots – 20% of its network today runs on fully open RAN technology with a true multi-vendor architecture, where third-party radios and cloud infrastructure come from different suppliers. The operator has integrated a modular SMO layer and is actively building rApps on top of it.
TELUS has partnered with P.I. Works as a key enabler of its open RAN journey, leveraging P.I. Works’ automation and intelligence capabilities to drive its evolution toward a fully open, multi-vendor architecture.
“We want to make sure that we have the best of breed and have the best solution coming in from different suppliers,” said Rawat. “We are willing to take the pain of integration between multiple parties because, at the end, it brings value to our end customer.”
Vendor dependency and operator-owned intelligence
A central theme of the conversation was the risk of relying on equipment vendors for network intelligence. Siradjev framed it plainly: “In terms of a silo, if your equipment vendor is controlling your intelligence in the network, then they are selling your decisions, as well as equipment.”
P.I. Works’ approach is to provide an intelligence platform that is fully separated from the equipment layer – open and vendor-agnostic – allowing operators to host applications from third parties, hyperscalers, or their own development teams.
“You can layer AI agents on top of this platform and generate bigger value for the end customer,” Siradjev said. “You can enable more revenue-generating use cases – for example, slice assurance in radio networks.”
Rawat pushed back on the assumption that RAN vendors withhold data. The issue, he said, has been the tooling – not the data itself. With Open RAN, TELUS now has contractual and architectural access to interfaces that were previously locked. The operator is building its own tooling to extract and process data across multiple layers – from bare-metal server sensors tracking CPU and memory, through the cloud and container layer, up to the SMO.
“Now you have multi-layered visibility into your system. You can do predictive analysis and preventive maintenance. You can cure a problem before it actually happens.”
5G SA, revenue opportunities today, and AI RAN promise
Both Rawat and Siradjev cautioned against waiting for 6G to start capturing value from intelligent networks. The tools and data to do so exist today in 5G SA.
“There is already an enormous amount of data being exposed,” said Siradjev. “The full concept of a disaggregated network already has enough information and there are a lot of use cases that can be built.”
He pointed to per-slice quality metrics in 5G SA as a foundation for new B2B, B2G, and B2B2C services – opportunities for operators to move beyond being a connectivity pipe and toward managed service providers.
On the relationship between Open RAN and AI RAN, both Rawat and Siradjev were direct: one does not work without the other.
“Open RAN is certainly a prerequisite for AI RAN,” said Rawat. “If it is not open, then you cannot control your AI. You are relying on somebody else’s intelligence.”
Siradjev described AI RAN not as a new technology, but as a continuation of what Open RAN set out to achieve – first through disaggregation enabling third-party applications, and now through AI-driven resource management and intelligent scheduling.
Proven today, built for what’s next
P.I. Works backs its vision with proven, commercially deployed solutions. Its EVO platform and suite of EXA AI-driven applications – spanning network automation, anomaly detection, and root cause analysis – are already managing complex radio environments for operators around the world.
As the industry moves toward AI RAN, P.I. Works’ position is clear: the operator, not the hardware vendor, should be the architect of network intelligence. And with the tools to prove it, that case is getting harder to argue against.
