Skip to content
18 Nov, 2026

FYUZ 2026 Recap: Agentic RAN, rApp Portability and Trust

Industry Telecoms

The path to network autonomy is no longer limited by what AI models can do. At FYUZ 2026 in Seattle, our conversations with operators, researchers and partners confirmed a clear shift: the industry has moved on from asking “can AI help?” to asking “how fast can I build an rApp, and can I trust it?”

That question sits at the heart of what Zinkworks does. Operators need agentic capabilities backed by a strong trust framework, and that is exactly what we brought to Seattle.

Key takeaways from FYUZ 2026

  • Agentic RAN is ready, if it is governed. Zinkworks launched rApp Studio Agentic Edition with live agentic and predictive demos at Stand 14.
  • Agents will orchestrate rApps, not replace them. Zinkworks and Vodafone shared how operators can scale from single purpose rApps to a managed rApp portfolio.
  • Portability is now a priority. Operators want rApps that move across SMOs rather than being tied to a single vendor.
  • Trust is the gating factor for autonomy. Confidence in AI decisions determines how quickly operators can climb the autonomous networks ladder.
  • New themes are rising fast. Integrated sensing and communications (ISAC), 6G and quality of experience (QoE) featured far more than expected.

FYUZ 2026: Launching rApp Studio Agentic Edition

During the show, Zinkworks launched rApp Studio Agent Edition, an intent-led agent that runs RAN operations end to end: monitoring, root cause analysis, change management, implementation and measurement. Agent’s reason over the RAN datalake but act only through certified rApps, under operator policy. The certified O-RAN control path is never touched, and every action is written to an immutable, intent-traced audit trail.

To show it working live rather than in a slide deck visitors saw two complementary demos:

  • Agentic demo: where the industry is heading. AI agents assessed network conditions and coordinated rApps within operator-defined policies. It drew the most discussion on the floor and opened many new conversations.
  • Predictive demo: value operators can capture today. AI forecast network behaviour and acted before issues reached customers. For teams closer to a buying decision, it was the clearest view of near-term return.

Together, the two demos showed how rApp Studio supports both stages on one platform: the future that agentic AI promises, and the measurable value predictive optimisation delivers today.

On Stage: From Single rApps to a Governed Portfolio

Signal Studio: Agents will Orchestrate rApps, Not Replace Them

James McNamara of Zinkworks and Francisco Martín Pignatelli of Vodafone explored how operators can scale from single-purpose rApps to a managed rApp portfolio.

Their central message: agents will not replace rApps; they will orchestrate them. rApps remain the trusted, testable building blocks of RAN automation. Agents add the coordination layer that decides which rApps act, when, and in what order. That model lets operators adopt agentic AI without discarding the investment and validation already built into their rApps.

Summit Stage: Recalibrating Open RAN Expectations

Zinkworks Solution Architect Michael Behan joined the panel “Towards rApp Portability and the SMO-RIC Reality,” part of the Summit Stage session “OpenRAN: Recalibrating Expectations, Redirecting Progress.” It was one of the best attended sessions of the week.

The panel looked at where Open RAN stands today and what it will take to keep it moving. Michael stressed that operators must build trust and confidence in AI as they climb the autonomous networks ladder. He closed the panel with insights from the TM Forum workshop Zinkworks co-hosted in London, explaining why platforms like rApp Studio are critical to that journey.

Shaping rApp Portability with Telecom Infra Project and AT&T

Zinkworks joined fellow members of the Telecom Infra Project (TIP) RAN Intelligence and Automation (RIA) group for a tour of AT&T’s lab in Redmond, led by AT&T Lab Director Michael McIntosh. The lab replicates every hardware and software component AT&T uses in the field. Each one is tested and validated for functionality, security and resilience before it can go live anywhere in the network.

The RIA group then reviewed the past 12 months of work and set its objectives for 2027. Zinkworks contributed ideas on rApp intelligence, rApp portability and interface standardisation, the same priorities operators raised with us throughout the week.

What Operators Told us at FYUZ 2026

Four themes came through consistently across stand conversations, panels and meetings.

  1. Agentic, AI-native operations are now the default conversation. Every operator discussion touched on agents. The difference now is a focus on control, auditability and safe execution.
  2. rApp portability across SMOs is a requirement. Operators want rApps that run across different Service Management and Orchestration (SMO) platforms, avoiding lock-in to a single vendor.
  3. ISAC and 6G are arriving sooner in planning cycles. Both topics came up far more often than we expected, signalling that operators are already thinking about how today’s automation choices will carry into next-generation networks.
  4. QoE at network speed is a growing challenge. In one session, Meta, YouTube, Axiata and Orange discussed how hard it is for operators to manage customer experience at the speed modern applications demand.

Our Take

The question has changed. It is no longer whether AI can help run the RAN. It is how quickly operators can build an rApp, and whether they can trust it in a live network. rApp Studio is designed precisely for that point: faster rApp creation, agentic orchestration and a governance framework that keeps every action accountable.

Start with our governed autonomy series

Our four-part series sets out the thinking behind rApp Studio Agentic Edition and the trust framework operators need for agentic RAN.

Let’s continue the conversation

Whether you are scaling from your first rApps to a managed portfolio, or preparing your network for governed agentic operations, Zinkworks can help you move faster with confidence.

Book a rApp Studio demo to see the agentic and predictive closed loops in action and discover how quickly your teams can build rApps you can trust.

FAQs

01 What did Zinkworks launch at FYUZ 2026?

Zinkworks launched rApp Studio Agentic Edition at FYUZ 2026 in Seattle. It enables operators to build and run agentic rApps for RAN optimisation within a governed trust framework, demonstrated through three live closed loops on a single architecture.

02 What is agentic RAN optimisation?
03 Will AI agents replace rApps?
04 Why does rApp portability matter to operators?
05 How does governed AI support the path to L4 autonomous networks?