Helping telcos scale agentic AI with control and speed

Ilan Sade is Division President, GenAI & Data at Amdocs. He outlines how Amdocs’ aOS aims to accelerate the shift to AI-native operations by connecting agents, people and workflows across existing systems.

TM Forum: Amdocs is focusing on its agentic operating system at DTW Ignite 2026. Can you briefly explain what differentiates aOS from other agentic AI approaches we’re seeing across the industry today?

IS: What differentiates aOS is that it is not just about building smarter agents, it is about how those agents, together with the right people, operate across real telecom environments.

First, aOS is purpose-built for telecom. It reflects deep domain expertise across care, billing, sales, and network operations, which allows agents to work within real telco processes rather than generic workflows. Crucially, it is designed for a hybrid operating model where AI and humans collaborate seamlessly across end-to-end business flows, rather than being confined to individual tasks.

Second, it is designed for ease of adoption. aOS works on top of existing BSS/OSS systems and across any AI or cloud infrastructure, so operators can integrate it into their current environment without needing to replace what they already have.

Third, it delivers faster time to value through pre-built, field-proven agentic capabilities and telco workflows. Operators can start with high-impact use cases and scale from there, rather than building from scratch.

Finally, it is open and modular. aOS is designed to plug into each operator’s agentic journey, allowing them to adopt capabilities incrementally and expand over time.

Together, this means operators are not just experimenting with AI, they can operationalize it as part of how the business runs, with AI and humans working together across systems and processes.

TM Forum: Bearing in mind the event’s motto, “The Future. Faster”, what needs to be in place for operators to deploy AI agents safely at scale, and how does aOS address that?

IS: To move faster with AI at scale, operators need more than capabilities, they need control and coordination across how work actually gets done.

In telecom, AI agents are operating across critical systems and customer-facing processes, often alongside human users. That makes consistency, oversight, and accountability essential – not just at the individual agent level, but across end-to-end workflows. Without that, scaling AI introduces fragmentation and risk rather than value.

aOS addresses this by providing a governance layer for how agents operate within these broader workflows. It applies guardrails and policies, ensures decisions are aligned with business rules, and enables coordinated interactions between AI-driven and human-driven steps. It also provides full traceability down to the data source.

This means operators can understand what actions were taken, why they were taken, and how they were executed across systems and roles.

That level of control is what allows AI to scale safely, supporting a new way of operating where AI and humans work together with speed, without compromising reliability or trust.

TM Forum: Can you explain how your approach aligns with TM Forum’s mission to accelerate the telco journey towards AI native?

IS: We designed aOS with one clear mission: to accelerate the telco journey to becoming an AI native and, ultimately, agentic organization.

That means not just embedding AI into operations but also making it practical for operators to adopt and scale over time.

A key part of our approach is openness. aOS is built to work with industry standards and existing BSS/OSS environments, which is critical in complex, multi-vendor telco ecosystems.

We also focus heavily on flexibility. Operators are at different stages in their journey, so aOS allows them to adopt specific capabilities based on their current priorities, whether that is improving customer care, enabling more dynamic sales interactions, or starting to connect workflows across domains.

Just as important is lowering the entry barrier. We address areas like security, data privacy, and integration upfront, so operators can move faster without needing to solve these challenges from scratch.

All of this is designed to support a gradual but clear shift. Today, many deployments focus on targeted use cases. Over time, those evolve into more connected, orchestrated workflows across business, IT, and network operations. That progression is what it means in practice to become AI native, and that is exactly what aOS is built to enable.

TM Forum: Where are you seeing the most immediate impact from aOS today in telecom environments?

IS: Today, the most immediate impact is coming from targeted, high-value use cases rather than full end-to-end transformation.

In customer care, AI is already improving how agents resolve issues, often working alongside human representatives to augment decision-making and accelerate resolution. In sales, we are seeing early traction with AI supporting more dynamic, personalized interactions.

These are areas where operators can move quickly and see clear returns, so that is where most deployments are focused today.

At the same time, the way these use cases are being deployed is already evolving. Rather than treating them as standalone capabilities, operators are beginning to connect them into broader workflows where AI and human actions are coordinated across domains such as care, sales, billing, and network.

That shift toward more connected, hybrid workflows is important. It reflects a move from isolated use cases to a more integrated operating model, which is where the next phase of value will come from.

TM Forum: How does aOS fit into the broader AI ecosystem operators are building with hyperscalers and partners?

IS: Operators are building their AI ecosystems with hyperscalers, models, and a range of tools, and that is the right approach.

Our role is to make that ecosystem work inside telecom operations. aOS sits on top and connects those capabilities to practical workflows, systems, and data.

For example, an operator may use a hyperscaler’s models for reasoning and analytics. aOS ensures those capabilities are applied consistently across processes like care or network operations, and those actions are accurately executed in the backend systems.

We are not replacing the ecosystem. We are enabling it to deliver outcomes in an operational context, where reliability and scale are critical.