In today’s hyper-connected world, the demand for uninterrupted telecommunication services is higher than ever. Telco networks emerge as the linchpin of modern connectivity, facilitating seamless communication and data exchange. But their effectiveness depends on their ability to operate without interruptions. Yet behind the scenes, many operators still rely on outdated, rule-based fault management systems that struggle to handle the complexity of modern networks.
These systems work well for straightforward, well-understood problems, but they fall short when it comes to new network technologies that don’t follow existing patterns, cascading failures with no clear entry point, or complex nonlinear alarm relationships. As a result, identifying the root cause of an outage can take hours or even days, leading to service impacts, dissatisfied customers, and reduced business success.
To address these challenges, Communication Service Providers (CSPs) are increasingly turning to Artificial Intelligence (AI) as a solution for smarter fault management. These AI-driven systems, fueled by vast volumes of alarm data, can identify root causes in real-time, even in complex, multi-domain environments, accelerating fault resolution while uncovering hidden patterns. It’s within this context that NOSSIS Genius ARISE, our new AI-powered fault management agent, emerges as a game-changer.
Meet NOSSIS Genius ARISE
NOSSIS Genius ARISE (Autonomous Recursive Intelligent System) is a groundbreaking AI Agent, part of the NOSSIS Genius (Altice Labs’ agentic AI framework for OSS), designed to revolutionize fault detection and Root Cause Analysis (RCA) in telco operations. It leverages advanced AI and Machine Learning (ML) models to bring operators new levels of real-time multi-domain RCA, scalability, and autonomy that were never available before, right out of the box.
Rather than relying on pre-written rules or static topology maps, it learns directly from the network itself, analyzing patterns across alarms to autonomously identify the fundamental origins of network issues, rendering it topology-agnostic and independent of prior network structure knowledge.
NOSSIS Genius ARISE addresses the inherent limitations of traditional rule-based systems, which often struggle with nonlinear and complex relationships. To overcome these challenges, the agent was developed to autonomously learn network behavior and uncover causality patterns based purely on relational dynamics, enabling the system to provide real-time root cause detection as patterns re-emerge.
Already gaining recognition within the industry, NOSSIS Genius ARISE won the AI Breakthrough 2025 Award for “AIOps Solution of the Year” and was also named a finalist in the “Network AI” category at FutureNet MENA 2025, underscoring its innovative approach and growing impact on the future of autonomous network operations.
How NOSSIS Genius ARISE works
At the core of NOSSIS Genius ARISE is a powerful AI-driven engine composed of three tightly integrated components:
Causality Matrix Generation
It employs a causality-driven preprocessing phase, which converts unstructured data into causality matrices, isolating meaningful patterns across time and domains.- Explainable Machine Learning Models
An advanced ML pipeline utilizes ML models to identify alarm clusters, feeding their output through a GenAI model to pinpoint root causes within these clusters, thereby combining pattern discovery with actionable insights. This multi-model approach creates distinct models for each predictor pair and automatically selects the best-performing rule to ensure adaptability. - Reinforcement Learning from Human Feedback (RLHF)
The system learns from feedback, refining its understanding and improving with each iteration to become more accurate and context-aware over time. Engineers can validate or correct root cause suggestions, further training the system to become smarter and more precise.
Together, this architecture allows NOSSIS Genius ARISE to operate effectively without any predefined network inventory or topology. It learns patterns autonomously and adapts as the network evolves. This is particularly valuable in environments introducing new network technologies or for operators with limited historical network knowledge, achieving significant alarm reduction. Its adaptability to evolving network conditions ensures its continued relevance and effectiveness in both mature and emerging networks.
This technology establishes new benchmarks in operational efficiency and fault management, paving the way for smarter and more autonomous operations. Demonstrating exceptional efficacy not only in established networks but also in new and evolving networks, where its AI-driven learning capabilities can surpass traditional human expertise in terms of speed and accuracy.
Additionally, beyond its core fault detection and RCA engine, NOSSIS Genius ARISE introduces a suite of advanced capabilities that elevate telecom operations to a new level of intelligence and autonomy. These features are designed not only to enhance operational efficiency but also to future-proof network management in increasingly complex and layered environments, including capabilities such as:
- Support for zero-touch NOC operations – Enabling new levels of autonomous network management.
- Cloud-native architecture – Deployable across public or private clouds, with support for microservice scalability and seamless integration with existing systems.
- Event- and time-agnostic detection – Identifying short- and long-term fault patterns, even across multiple domains and technologies.
- Business- and technology-agnostic adaptability – Allowing continuous alignment with topology changes and supporting multi-vendor, multi-domain environments.
- Customizable learning parameters – Allowing for the tuning of model confidence levels and other key thresholds.
- Integration with omnichannel virtual assistants – Enhancing customer experience and operational efficiency.
Its business and technology-agnostic design ensure compatibility across multi-vendor ecosystems and evolving topologies, while customizable learning parameters allow operators to fine-tune performance based on their unique operational needs.
Real-world impacts and benefits
This isn’t just automation, it’s autonomous intelligence. NOSSIS Genius ARISE delivers meaningful real-world benefits for operators across multiple continents, bringing both operational and customer experience improvements, including:
- 26% average alarm reduction on top of a rule-based mature operation.
- 33% fewer tickets by clustering alarms and avoiding duplicate trouble tickets (TTKs).
- Accelerated root-cause detection and resolution, in real use-cases, it was able to anticipate the detection by 7 hours.
- Accurate detection in cascading failure scenarios.
- Fast adaptation to new environments, autonomously extracting meaningful correlations after the first month.
In its real-world deployment, the system has demonstrated extensive operational effectiveness across multiple telecommunications operators on three continents, demonstrating its broad applicability.
Final Takeaways
As networks grow more complex and less predictable, traditional systems are no longer sufficient to handle the dynamic nature of modern telecom environments. The need for intelligent, adaptive fault management becomes not just a technical requirement but the linchpin for telco success. NOSSIS Genius ARISE steps into this gap with a transformative approach, combining causality-driven AI, explainable ML, and reinforcement learning to deliver fault detection and resolution that is faster, smarter, and more autonomous.
Combined with the broader suite of NOSSIS Genius agents, it lays the foundation for telcos on the road to fully autonomous operations, where intelligence, agility, and resilience define the next generation of network management.
Ultimately, NOSSIS Genius ARISE delivers a multi-metric gain, improving service reliability, operational knowledge, and cost efficiency, representing a transformative solution that aligns business goals with technological innovation to drive excellence across telecommunications operations.