Case study overview
- This case study explores stc’s journey toward autonomous network operations, driven by its investment in a best-in-class, cloud-native telco platform and a strategy focused on AI-powered automation and operational efficiency.
- It highlights how stc is implementing a modular, AI-enabled framework aligned to the OODA loop, significantly reducing manual effort and enabling intelligent observability, decision-making, and remediation across its cloud infrastructure.
- Engaging STL Partners, stc and Cisco were keen to share key enablers, outcomes, and learnings from building its telco cloud transformation, showcasing how AI, automation, and agile practices can transform telecom operations at scale.
Introduction
The telecommunications industry is undergoing a profound transformation driven by the convergence of cloud-native architectures and cloud-native practices, supported by the growing maturity of artificial intelligence (AI).
stc is at the forefront of this shift towards zero-touch network operations, through investing in a building a best-in-class, multi-vendor telco cloud platform. To advance its journey towards autonomous networks, stc is pursuing a dual strategy:
- Drive operational efficiencies by enhancing internal network automation through AI (e.g., faster remediation, cost savings)
- Ensure efficient use of resources through granular cost allocation and, through this, create additional value by offering cloud infrastructure and intelligent services beyond its core internal customer base
This case study explores how stc is turning this vision into reality—detailing the collaborative approach, key enablers, and strategic milestones on its path toward autonomous network operations.
The vision for delivering autonomous networks
Realising the vision of autonomous networks requires more than introducing automation technologies — it demands intelligent observability delivered through a single view of system health, that can analyse anomalies in real time and support more efficient decision making. Autonomous networking capabilities are built around four key pillars that align with the OODA loop (Observe – Orient – Decide – Act).
Figure 1: The four key pillars of autonomous networking capabilities
- The data collection layer (observe): The entire telco cloud infrastructure observability data — including syslog, performance, and fault management — for compute, networking, platform and management systems is collected and centralized in a single AIOps platform, enabling a single- plane of glass visualization and data processing.
- The data composition layer (orient): The architecture leverages AIOps functionality to perform data correlation, normalization, deduplication, and visualization to derive data insights across multiple platforms. AI is also leveraged for observability through automated baselining, anomaly detection, noise reduction and KPI forecasting to support sound decision making.
- The decision execution later (decide): This stage is evolving from rules-based, human-led control to an AI-assisted model — where AI-generated hypotheses are used to identify root causes using RAG-based methods. Semi-automated decision-making workflows, supported by human-in-the-loop oversight, enable the validation of these hypotheses. This approach maintains accountability and ensures safeguards are in place as AI adoption scales.
- The process execution layer (act): Actioning a remediation relies on a complete stack of network, workload and test automation. This enables a range of use-cases including automated data center E2E deployment, configuration change, data center expansion, hardware replacement, new workload onboarding and workload lifecycle management. This provides a solid foundation for future remediation processes that are triggered by AI yet remain under human supervision.
Of these pillars, Detect and Act are currently the most mature, while Analyse and Decide are still being refined through ongoing pilots and iterative testing.
Implementing the above framework has resulted in a reduction of manual work by up to 50% for a new data center deployment, 50% for new workload onboarding and up to 75% for executing automated testing.
Figure 2: The KPIs for measuring the success of autonomous networks
The building blocks for success
To support its journey towards autonomous operations, stc and Cisco have focused on key enablers:
- Cloud-native architecture: A robust, cloud-native architecture, that is aligned to NVIDIA frameworks, ensures network resilience. This is further ensured through additional data centre deployments.
- Automation-first principles: Integration into CI/CD pipelines and assurance systems supports faster iteration and reduced manual overhead.
- Data readiness and AIDataOps: High-quality structured data, coupled with predefined actions, form the foundation for effective AI model training and continuous refinement. A robust AIDataOps approach ensures that data pipelines are reliable, scalable, and governed—enabling consistent performance as AI use scales.
- Agile, DevOps-driven culture: Embracing agile methodologies and DevOps practices ensures continuous delivery, cross functional collaboration and responsiveness to ongoing change and modernisation.
- Modular AI-automation architecture: Ideal architecture structures automation into small ‘micro’ blocks, which improves transparency and auditability, while making each component easier for developers to assess. Developers can spend less time writing complex end-to-end scripts, and more time reviewing and validating smaller, well-scoped components – improving overall system reliability and enabling more rigorous human-in-the-loop oversight.
- People and skills: Engineers and developers that leverage AI to enhance skills and productivity, streamlining repetitive script writing and shifting focus towards reviewing and validating automated processes. This evolution improves operational efficiencies while also enabling engineers and developers to focus on architecture-level thinking and take on higher-value roles earlier in their careers.
- Standardization: Using industry standard frameworks such as API for data exchange, CI/CD for change management, robot framework for automation, Infrastructure-as-Code and YANG for network automation, as well as ETSI for workload onboarding and management.
Figure 3: Daily task distribution of developers, before and AI enabled
Vision and next steps
stc’s vision for autonomous networks is grounded in a commitment to continuous improvement and innovation. As the telco deepens its adoption of AI and automation, the next phase of its evolution will focus on refining the use of AI for autonomous test generation. By enabling predictive testing and accelerating validation cycles, stc aims to further enhance the reliability and agility of its network operations.
In parallel, automation will continue to be embedded more deeply across operational workflows, with expansion into new and increasingly complex use cases. This will support greater scalability and efficiency as the network becomes more self-governing.
Beyond internal gains, stc is also exploring ways to monetise its investments in cloud-native infrastructure and AI capabilities. By offering intelligent services and digital infrastructure solutions to external partners, it is positioning itself to unlock new revenue streams and extend the value of its transformation journey across the broader telecoms ecosystem.
