From Generic to Verticalized: Amdocs Unveils a Standard for Skilled, Trustworthy Telco-Grade AI Agents

In collaboration with Amazon Net Companies (AWS) and NVIDIA, whitepaper introduces a framework for telco-verticalized AI brokers, emphasizing ontology, reasoning, simulation, belief and model engineering

JERSEY CITY, NJ / ACCESS Newswire / June 16, 2025 / Amdocs (NASDAQ:DOX), a number one supplier of software program and providers to communications and media firms, as we speak unveiled a brand new normal defining what makes a telco AI agent a complicated, verticalized enterprise providing. In a whitepaper titled “AI Verticalization for Telcos,” Amdocs presents an overview for telco-grade brokers which are expert, brand-engineered, reliable, and autonomous-capable of reworking how CSPs have interaction with prospects and function their networks.

Whereas the telecom sector has embraced generative AI at tempo, many deployments fall wanting true operational integration. This whitepaper defines a Telco-Grade Agent as one that’s “verticalized”-deeply embedded with telco-specific abilities, ontologies, and reasoning capabilities. These brokers transcend surface-level interactions, delivering network-aware, context-sensitive responses and executing clever decision-making rooted in area experience.

“Within the age of AI, the distinction between automation and intelligence lies in specialization,” mentioned Anthony Goonetilleke, Group President of Know-how and Head of Technique at Amdocs. “Telco-grade brokers aren’t simply smart-they’re industry-native. They converse the language of networks, providers, and prospects, and so they act with objective. This framework ensures that CSPs don’t simply implement AI-they implement it proper.”

Key facets of a telco-specific AI agent embrace:

Ontology is the Spine of Telco Intelligence

For AI brokers to function successfully within the telecommunications {industry}, they need to perceive the language, logic, and construction of telco methods. That’s the place ontology – a proper illustration of area data that permits constant information interpretation throughout platforms and duties – turns into vital. A well-defined telco ontology supplies AI brokers with a contextual understanding of service plans, technical specs, billing buildings, and buyer interactions. To assist this, a strong ingestion pipeline is needed-capable of dealing with a wide range of information varieties that brokers depend on, together with real-time streams, time sequence information, and different structured and unstructured sources.

By leveraging in-context studying and telco-specialized giant language fashions (LLMs), brokers acquire superior reasoning capabilities and adaptableness. This structured framework ensures AI brokers ship correct, related, and uniform responses, whereas permitting for seamless collaboration throughout methods and sub-agents.

Autonomy and Predictive Reasoning That Works

AI brokers should function with true autonomy-detecting points, initiating actions, collaborating with different brokers, and escalating to people when needed-all earlier than a buyer ever reaches out. This shift from reactive assist to proactive service unlocks measurable features in effectivity and buyer satisfaction.

Past entry to telco information, brokers should be capable to purpose with it-applying domain-specific logic, workflows, and greatest practices to ship choices and actions aligned with {industry} expectations. That is what permits actually clever, telco-effective autonomy.

Digital Twin and Simulation

For operators, a digital twin is a real-time digital reproduction of their bodily community, designed to assist dependable, data-driven decision-making. By constantly ingesting information and enabling high-fidelity simulations, it supplies actionable insights that result in trusted, environment friendly execution. Information, analytics, and AI/ML type the core basis of any efficient digital twin.

Simulation performs a essential position by providing a secure and scalable setting to coach and take a look at AI brokers. It permits agentic AI to study complicated decision-making, anticipate outcomes, adapt methods, and function autonomously-without real-world danger. When mixed with a telecom community digital twin-fed in actual time by information from base stations, routers, and consumer devices-this creates an autonomous community that may self-optimize, dynamically handle sources, and even self-heal. The end result: improved efficiency, decrease operational prices, and a considerably enhanced buyer expertise.

Model Engineering: Designing the Digital Face of the Enterprise

As AI brokers change into the primary touchpoint between firms and their customers-across industries, not simply in telecom-how they sound, look, and reply carries actual model weight. Brokers can’t simply be generic; they need to embody the voice, tone, and persona of the model in motion.

This shift issues as a result of prospects are paying consideration. In a global study* commissioned by Amdocs in collaboration with McCann Tech company and Coleman Parkes, 80% of shoppers mentioned they belief AI brokers to resolve service points, and 60% imagine AI can positively form their notion of a brand-when performed proper. That belief isn’t automated; it’s earned via intentional design, from the agent’s voice and visible presence to the way in which it behaves and adapts to every interplay.

Belief Can’t Be Trivial

Enterprise-grade Brokers should additionally possess strong non-functional capabilities. These horizontal capabilities embody essential facets like belief and safety, that are elementary to the reliability and effectiveness of GenAI options within the telecommunications sector. By integrating these horizontal attributes, telco brokers not solely improve their service supply but in addition make sure that operations are safe, compliant, and reliable, thereby fostering stronger buyer relationships and confidence.

Ongoing Collaboration with Amazon Net Companies (AWS) and NVIDIA

Amdocs continues to work intently with NVIDIA to combine NVIDIA’s full-stack AI foundry service – a group of NVIDIA AI Basis Fashions, NVIDIA NeMo™ framework and instruments, and NVIDIA DGX™ Cloud that offers enterprises an end-to-end answer for creating customized generative AI fashions – with Amdocs’ generative AI providing, amAIz Suite.

The collaboration has targeted on harnessing generative AI to remodel buyer experiences, community automation and ship operational efficiencies. Amdocs leveraged NVIDIA DGX™ Cloud on AWS to check and tune a number of fashions, whereas utilizing NVIDIA NIM microservices, part of the NVIDIA AI Enterprise software program platform, to quickly deploy fashions for seamless, scalable inference with industry-standard APIs.

Amdocs has partnered with AWS to construct enterprise-grade, cloud-native AI options that ship production-ready capabilities for communication service suppliers (CSPs). By this collaboration, Amdocs’ amAIz Suite integrates the excellent generative AI stack of AWS-including Amazon Bedrock for basis mannequin entry, AWS purpose-built information providers for real-time processing, and scalable cloud infrastructure-to allow telco-grade brokers that meet the essential necessities for enterprise AI: adaptability, scalability, information integration, and adherence to privateness and safety requirements.

Amdocs, AWS, and NVIDIA have demonstrated a number of collaborations that present the real-world impression of GenAI in telecom. The primary, introduced at Digital Transformation World Asia, explored how GenAI can improve billing and buyer care and gross sales experiences. The second, unveiled at Cell World Congress Barcelona and NVIDIA GTC, launched a framework for autonomous community operations via clever community brokers.

Amdocs might be demonstrating its joint AI agent options with NVIDIA at TM Forum’s Digital Transformation World Ignite in Copenhagen, June 17-19, 2025.

*Survey, interviews and focus teams of 120 CSP leaders and seven,025 shoppers aged 18-69 in 14 international locations throughout North America, Europe, and Asia.

Supporting Assets

About Amdocs

Amdocs helps those that construct the long run to make it wonderful. With our market-leading portfolio of software program services, we unlock our prospects’ modern potential, empowering them to offer next-generation communication and media experiences for each the person finish consumer and enterprise prospects. Our staff across the globe are right here to speed up service suppliers’ migration to the cloud, allow them to distinguish within the 5G period, and digitalize and automate their operations. Listed on the NASDAQ International Choose Market, Amdocs had income of $5.00 billion in fiscal 2024. For extra info, go to http://www.amdocs.com

Amdocs’ Ahead-Trying Assertion

This press launch contains info that constitutes forward-looking statements made pursuant to the secure harbor provision of the Personal Securities Litigation Reform Act of 1995, together with statements about Amdocs’ development and enterprise ends in future quarters and years. Though we imagine the expectations mirrored in such forward-looking statements are based mostly upon affordable assumptions, we may give no assurance that our expectations might be obtained or that any deviations is not going to be materials. Such statements contain dangers and uncertainties that will trigger future outcomes to vary from these anticipated. These dangers embrace, however will not be restricted to, the results of basic macroeconomic situations, prevailing degree of macroeconomic, enterprise and operational uncertainty, together with on account of geopolitical occasions or different regional occasions or pandemics, modifications to commerce insurance policies together with tariffs and commerce restrictions, in addition to the present inflationary setting, and the results of those situations on the Firm’s prospects’ companies and ranges of enterprise exercise, together with the impact of the present financial uncertainty and {industry} stress on the spending choices of the Firm’s prospects. Amdocs’ capability to develop within the enterprise markets that it serves, Amdocs’ capability to efficiently combine acquired companies, hostile results of market competitors, speedy technological shifts that will render the Firm’s services out of date, safety incidents, together with breaches and cyberattacks to our methods and networks and people of our companions or prospects, potential lack of a serious buyer, our capability to develop long-term relationships with our prospects, our capability to efficiently and successfully implement synthetic intelligence and Generative AI within the Firm’s choices and operations, and dangers related to working companies within the worldwide market. Amdocs could elect to replace these forward-looking statements in some unspecified time in the future sooner or later; nonetheless, Amdocs particularly disclaims any obligation to take action. These and different dangers are mentioned at better size in Amdocs’ filings with the Securities and Trade Fee, together with in our Annual Report on Kind 20-F for the fiscal yr ended September 30, 2024, filed on December 17, 2024, and our Kind 6-Ok furnished for the primary quarter of fiscal 2025 on February 18, 2025, and for the second quarter of fiscal 2025 on Could 19, 2025.

Media Contacts

Michael Zema
Amdocs Public Relations
[email protected]

SOURCE: Amdocs Administration Restricted