Rakuten Symphony on Balancing AI Agent Autonomy with Network Control

Why It Matters

Policy‑driven AI agents enable telecoms to automate complex operations safely, cutting costs while enhancing service reliability and customer satisfaction.

Key Takeaways

• AI agents must operate under policy‑as‑code guardrails
• Rakuten Symphony treats the network as a software platform
• Agent‑to‑agent protocols replace manual API chaining
• Zero‑touch operations reduce OPEX and accelerate issue resolution
• Context‑aware agents improve customer experience and decision‑making

Summary

The discussion centered on how generative AI agents are reshaping telecom operations, with Tapas Ranjan of Rakuten Symphony arguing that their net benefit hinges on policy‑based execution rather than unrestricted autonomy. He described the industry’s "control paradox"—the need to grant agents freedom while preventing non‑deterministic behavior—and proposed embedding policies as code to create a safety envelope around AI actions.

Ranjan explained Rakuten Symphony’s evolution from traditional hardware‑centric networks to a fully virtualized, software‑defined platform that supports an "agent‑aware" ecosystem. By moving engineers from manual API chaining to agent‑to‑agent protocols, the company enables domain‑specific agents to consume service catalogs, correlate tickets, changes, and alarms, and execute remediation tasks autonomously. This shift underpins their zero‑touch operations roadmap, promising reduced operational expenditures and faster time‑to‑market.

He highlighted concrete use cases: an AI agent diagnosing whether a network issue stems from a recent configuration change, and a context‑aware agent delivering personalized customer support while staying within predefined guardrails. These examples illustrate how policy‑driven AI can act as an advisor and orchestrator without granting "the keys to the kingdom," preserving vendor‑agnostic flexibility and preventing lock‑in.

The implications are clear: telecom operators that adopt policy‑as‑code frameworks and fully virtualized network APIs can unlock AI‑driven efficiencies, lower costs, and deliver superior user experiences. Rakuten Symphony’s approach offers a blueprint for the industry to balance autonomy with control, accelerating the transition to truly autonomous, zero‑touch networks.

Original Description

Tapas Ranjan, VP of applied AI at Rakuten Symphony, explains how AI agents represent a net positive for telecom when implemented with policy-based execution frameworks. He discusses the shift from manual API management to agent-driven automation and shares how Rakuten Symphony has built an agent-aware ecosystem using domain-specific agents and service catalogues.

Featuring: Tapas Ranjan, Vice President, Applied AI, Rakuten Symphony

Recorded June 2026

#NetworkAPIs #AIAgents #TelecomAutomation #ZeroTouch #CloudNative #TelecomTV