We salute the nine technology award winners and the independent panel of judges who had the unenviable job of choosing them from a pool of exceptional entries. Our congratulations and thanks to everyone who entered. Read on to discover more about each of the award-winning entries.

AI-driven Orchestration Award

This Award is for the most innovative automated service orchestration solution powered by AI from a vendor. Amdocs Service & Network Orchestration is the winner, powered by the company’s aOS – agentic operating system. Amdocs proved that agentic AI-driven orchestration is delivering value today through customer engagements, Catalyst proofs of concept and measurable key performance indicators (KPIs).

By fusing agentic AI, trusted intelligence and a real-time view of the network based on a digital twin, Amdocs Service & Network Orchestration converts fragmented network data into knowledge. This means communications service providers can move from reactive operations to automated, intent-driven orchestration.    

Agentic AI turns intent into outcome: the solution allows customers to express requirements in natural language or a visual format, such as, “secure, low-latency connectivity for a remote mine”. Intelligent agents decompose this into TMF701-compliant (a TM Forum REST API for process flow management) service definitions which reduce workflow creation time from 12 weeks to one hour.

Intent decomposition agents translate business needs into precise service and resource requirements; feasibility, policy and optimisation agents validate constraints across the RAN, transport, edge compute, slicing capacity and resilience budgets, using real-time network inventory, topology and digital twin simulations.

This ensures that every service design is safe, feasible, and executable, eliminating manual engineering loops and avoiding back-and-forth qualification cycles. This level of trusted autonomy is foundational to achieving reliable Level 4 behaviour, allowing downstream orchestration to execute with zero-touch confidence.

Agentic fallout management as part of the fulfilment process is one example of where the solution can be a major differentiator. It delivered: 20% to 40% fewer missed milestones; 20% to 30% faster resolution time; 40% less manual intervention; plus more than 95% automation of fulfilment tasks.

Similarly impressive results in closed loop assurance for autonomous operations (using various TM Forum open APIs) and business impact enable or should enable monetisation of connectivity including 5.5G, 6G, network slicing, enterprise private networks and edge‑driven services.

AIOps Award

This Award goes to Reailize for the best operations solution from a vendor that incorporates AI functionality in the shape of TelcoGPT – an Agentic AI framework, purpose-built for telecom diagnostics. It is used in production by two Tier-1 operators and two lab deployments in the US, and has extensive trials with three European Tier-1 operators.

When one of their clients experienced a major outage, a large engineering team manually reviewed packet traces, alarms and logs without success. TelcoGPT analysed the same incident data, identified the root cause within minutes, and recommended the appropriate remediation.

TelcoGPT’s breakthrough architecture combines deterministic telecom intelligence with agentic AI reasoning. Independent evaluations show 92% accuracy with compact models (OSS-120B), scaling to near 100% when integrated with frontier models like Gemini.

The framework delivers value across three key areas:

• Level 1 operations support –  triage correlates tickets, traces, logs and network data to identify the domain where the failure most likely occurred. It guides first-line engineers toward the correct resolution path, reduces operational noise and escalations to higher support tiers.
• Level 3 performance troubleshooting – analysing packet-level data alongside packet capture traces, telemetry and 3GPP specifications to reconstruct subscribers’ journeys, interpret protocols’ behaviour and analyses explainable root cause and mitigation steps. Otherwise complex failures can take days to diagnose.
• the Lab Agent automatically routes failures to the right teams, accelerating validation cycles and eliminating bottlenecks caused by manual triage in telecom labs.

During formal business case evaluations, Tier-1 operators recorded operational improvements such as a 67% reduction in mean time to resolve (MTTR) incidents, half the number of escalations to higher support tiers and time to diagnose complex failures fell 94%. The framework demonstrated a 17-fold increase in troubleshooting capacity and freed up 27,000 engineering hours annually.

Automation Solution Award

Ericsson’s Operation Engine won this award for the leading solution for network Automation by a vendor. The Company says that its AI-driven Autonomous Fiber Network Management solution is a key milestone in the Touch Free Automation journey. This solution integrates machine learning-driven simulation with automated root cause analysis to prioritize fiber incidents across the network. More importantly, it is already making a tangible impact – transforming fiber incident management and delivering more resilient, reliable outcomes for the True network.

The system automatically classifies severity, creates a structural urgency index, daily field action lists and zero‑touch ranking for decisioning in network operations centres. The prioritisation is autonomous and topology-aware, predicting structural topology bottlenecks before outages occur, and applying proactive assurance and autonomous risk forecasting.

The measurable impact between Q4-2025 and Q1 in 2026, across multiple nodes and regions, included: reduction in MTTR, over 90% automation of prioritisation decisions and more than 90% accuracy in predictions about nodes going down via simulations. The engine helped to improve customer experience by prioritising non-service impacting Fiber incident tickets to work and get them fixed by Field, and improved uptime through proactive elimination of double‑cut risks.

Ericsson says the Engine is vendor-agnostic, with a continuously evolving autonomous digital twin synchronised with live service and resource state. Machine learning (ML) assists the simulation of millions of potential failures and automatically correlates possible incidents and simulated impact. It offers a practical blueprint that can be deployed at scale over large networks.

Autonomous Network Operations Award

Netcracker and Swisscom received the Autonomous Network Operations Award. After working closely with Netcracker on its vision and target architecture, Swisscom decided the only way to scale automation companywide was to first break down its organisational structure and eliminate tight IT interdependencies. The operator’s approach, including the creation of a pioneering domain automation blueprint, offered as IT-as-a-Service, has set a global benchmark for autonomous operational excellence in telecom.

Swisscom reorganised into 30 small, separate domains, each self-sufficient with its own NetDevOps teams. Each had its own NetDevOps team which controls the domain’s IT stack and operates like a small business within Swisscom, leaning on the principles of TM Forum’s Open Digital Architecture. This gives each team full autonomy and speeds up the introduction of agentic AI.

IP transport was the first domain to be modernised and automated using Netcracker’s Network Domain Automation Solution. Service extensions were cut from six to nine months down to two weeks. Also there were: 50% fewer failures, 30% cost savings and a federated inventory that provides a key foundation for agentic AI.

Taking learnings from IP transport, Swisscom and Netcracker created a domain automation blueprint to replicate success by ensuring every domain adheres to a common set of industry standards, tools, assets, systems and best practices. In effect, the two companies offered this blueprint as a pioneering IT-as-a-Service to Swisscom’s internal teams, allowing each domain to modernize and run autonomously with the same software development and security environments, and about 80% of each domain has the same OSS components.

Six more domains in Swisscom followed, as well as cross-domains, with more to come. They represent the highest value domains to the business, operating at autonomy Level 4 and above. The new domains rolled out in around six months, down from nearly two years for the first one. Development costs fell by 40% through shared assets and cross-domain synergies.

“We needed to make it easier and more convenient to consume IT-as-a-Service than to build it themselves in their domain. The IT-as-a-Service developer experience is key to scaling the modernization and automation of new domains across the business,” noted Rudolf Strijkers, Lead Architect at Swisscom.

Autonomous Network Operator Leadership Award

Elisa came out on top in this Award category designed to recognise the most successful advances towards Level 5 network autonomy. Over the last 15 years, network complexity and data volumes have increased 22x — yet incidents have fallen by 81% and 99% of faults are resolved before customers experience them. This has resulted in 0 A-scale incidents during the last 9 years ultimately leading to a decade of continuously improving customer experience.

Building the foundation began in 2010 with a dedicated Service Operations Centre and the start of automated network operations. By 2012, network operations were fully automated; by 2015, Elisa had automated network optimisation.

The second phase, 2015–2024, saw the Elisa Defined Network (EDN) created in-house. It became the engine of Elisa’s AI-driven network planning and provisioning from 2019, then the base for cloud-native network automation from 2022. With the Digital Twin operational since 2018, Elisa has been able to simulate network changes before deploying to live network.

In 2025, the third, most ambitious phase commenced, combining the EDN platform and 15 years’ proprietary automation IP with Google Cloud’s Autonomous Network Operations framework. Elisa’s network is moving from predictive to prescriptive analytics, and from reactive to fully proactive management.

Already agents autonomously troubleshoot, validate configurations, detect anomalies and collaborate with human operators in real time, in live operational networks. The strategic cooperation spans the entire network – transport, core and radio.  The goal is to achieve world-leading network autonomy: Elisa is already operating at Level 4.

The next frontier is Level 5 – a self-managing network without human intervention. Elisa will expand autonomous operations across every process domain, from reactive fault management through planned network actions to fully autonomous planning, implementation and service delivery. It intends to have zero incidents in 2030.

Customer Experience Award

China Mobile and Huawei Technologies received the Customer Experience Award in recognition of their innovative use of AI to enhance customer experience in the vendor category.

According to China’s National Health Commission, two out of three people are vulnerable to medical misinformation. Unnecessary annual medical spending exceeds RMB 100 Billion, while more than 60% of doctors report spending significant time correcting patients’ health misconceptions.

To address this challenge, China Mobile and Huawei developed the AI Health Assistant, using AI Calling as a new healthcare service entry point. Built with the IMS Data Channel (IMS DC), the solution synchronizes voice communication with real-time data transmission, allowing users to access healthcare services simply by making a phone call. Patients can interact with the AI Health Assistant through natural conversation and interactive menu, without downloading an app, creating an account, remembering passwords, or navigating complex interfaces.

During the call, users can be presented with interactive menus and lightweight visual interfaces while continuing a normal voice conversation. The IMS DC layer supports mini-app-style service templates, although no app is required on the caller’s side. These templates can support appointment booking, insurance inquiries, medical guidance, and other workflows, enabling closed-loop services within the call itself. This dual-mode interaction — voice plus visual guidance — improves accessibility and clarity beyond traditional voice-only hotlines.

An AI service engine orchestrates the overall experience. Large language models interpret natural speech, understand user intent, generate healthcare guidance, and respond in a human-like voice. At the same time, the interactive interface displays suggestions and contextual information in real time.

An intelligent service server manages identity authentication, context continuity, and service orchestration, helping ensure secure, accurate, and personalized interactions.

The platform connects directly to the backend systems of ecosystem partners, including 5,000 hospitals and 1 million doctors. During calls, users can upload medical reports or photos of medication for real-time analysis, supporting a closed-loop service journey from consultation and report interpretation to registration guidance and appointment booking.

By making healthcare network-native, the solution reduces fragmentation across different platforms and makes digital health services more universally accessible — starting with a single phone call.

Network AI Award

Rakuten Symphony won this Award for its Rakuten AI for Telecom, judged to be the most innovative AI solution to enhance network performance by a vendor. Operators struggle to extract actionable value from their growing volumes of data: Rakuten AI for Telecom is designed to bridge the gap between telemetry and autonomous performance.

The platform, built on a scalable microservices architecture, has been live on Rakuten Mobile’s nationwide network in Japan for more than two years. It supports over 400 telecom experts and more than 120 applications. It integrates ML and Generative AI to access and correlate data from disparate domains – Network, IT and customer behaviour – handling 350 TB of data daily and managing a repository of 15 PB.

“At Rakuten Symphony, we recognized early on that the future of telecom isn’t just about connectivity—it’s about the intelligence layer governing it. This award reflects our commitment to a truly autonomous industry. We’ve designed our platform to act as the ‘brain’ of the network, where AI agents collaborate in real-time to optimize performance, resolve issues, and adapt to demands. We aren’t just building a product; we are building the foundation for the autonomous telco, ensuring operators can scale their performance and their ambition in equal measure.,” says Tapas Ranjan, Vice President, Product Management, Rakuten Symphony.

The AIOps framework monitors data lake components to extract immediate insights for proactive fault detection, which prevents subscriber downtime. The AI Hub stores, versions and deploys trained, telecom-specific models for rapid inferencing. The plan is to continuously evolve the network’s ‘brain’.

Rakuten Mobile says it is the first in the world to validate Level 4 autonomy with TM Forum for live RAN energy efficiency: the remote electrical tilt AI model has boosted spectral efficiency by 25% and increased handover rates from 97% to 99%, improving connectivity in high-density and underserved regions.

By correlating concurrent events, AI speeds up approvals and automates the resolution of recurring issues using playbooks and reducing resolution times by 60%. The Intelligence Site Audit cuts review times, delivering up to $1.6 million in annual savings; automated monitoring reduces total operational costs by 30%.

Finally, the ‘brownfield-first’ strategy means legacy operators can adopt these and the platform’s other innovations without a ‘rip and replace’ approach to facilitate a global shift toward more efficient, sustainable and intelligent infrastructure.

Network Data & Intelligence Innovation Award

P.I. Works and Turkcell won this category for the leading data and intelligence solution or platform transforming network operations in the vendor category. P.I. Works deployed the EXA Disaster Recovery Automation (DRA) solution across Turkcell’s nationwide network, covering multi-vendor nodes for all RAN technologies.

The near real-time automation maintains connectivity when it is needed most, such as during natural disasters or regional power failures: standard procedures often lack the necessary speed to prevent widespread outages which can endanger life. Managing sites manually is complex, slow and resource-intensive, especially when hundreds of locations are impacted simultaneously.

Resilient networks require more than traditional operational processes. Through AI-driven automation and predictive network intelligence, we are evolving toward a more proactive operational model that improves service continuity, optimizes energy usage, and enables faster response during critical events,” said Vehbi Çağrı Güngör, Chief Technology Officer at Turkcell

P.I. Works’ automated DRA module was integrated directly with data about performance management, call trace and fault management. The system automatically detects power cuts and immediately initiates power reduction or cell/transmitter-receiver locking, based on battery backup times.

The AI-driven forecasting engine dynamically adjusts distributed energy saving parameters, extending battery life by up to several hours, based on KPIs from vendors. This has contributed to sustainability goals and reduced operational costs.

P.I. Works’ Geolocated Customer Experience Solution applies energy saving to the most critical sites. Once power is restored and alarms are cleared, the system automatically reverts all parameters to their original states to resume normal operations.

The deployment has improved Turkcell’s operational efficiency and network reliability. The solution carries out 3000 cell deactivation/activation and 27000 parameter change total 30,000 closed-loop actions for 500 power outage alarms monthly, reducing the burden on network operations teams and managing disaster recovery processes to keep customers connected.

Operator Deployment Award

Deutsche Telekom was deemed to have the best example of an AI or automation deployment with this Award. Its RAN Guardian relies on multiple, specialised AI agents working together. Most telecom automation systems are rule-based whereas RAN Guardian uses AI agents that reason, plan and act, enabling more flexible and adaptive operations.

Before introducing RAN Guardian, operations teams could monitor and manage only around 1,000 events – such as sports fixtures or festivals – per year. They required manual configuration changes and continuous monitoring to stabilize the network, often taking more than an hour.

Since its launch in November 2025, RAN Guardian has reduced the time required to manage major events from hours to roughly one minute – an improvement of more than 95%.

It works by multiple specialised AI agents working together. The Event agent constantly scans event listings and public data, then locates and categorises events, and determines their size. The Monitoring agent takes the event data and evaluates the RAN’s capacity to handle the expected data load. It analyses whether neighbouring antennas can provide additional support, then monitors network parameters in real time, detects anomalies or congestion risks early on.

The Remediation agent implements optimisation measures, such as reallocating resources or adjusting configurations. All actions are saved so they can refine the agents’ future performance.

Here are some of the results achieved by the RAN Guardian:

• more than 100 network optimisations were triggered automatically in the first month
• AI identified 237,000 public events in 2026.
• in February ‘carnival’ season alone it monitored 130 large events.

Beyond events, the AI agent has had outstanding results in integrating fragmented data and accelerating troubleshooting in wider contexts. After its success in Germany, RAN Guardian is now scaling across DT’s other European markets, starting with the Czech Republic and Croatia.

Next month, we’ll publish an article highlighting the individuals who received awards at FutureNet World 2026.