Huawei Intelligent Operations Drives AI and Gen-AI Innovation in Telecom Networks

Huawei Intelligent Operations Advances AI and Gen-AI in ICT

Operating telecom networks is becoming more complex. Fixed and mobile networks often converge to deliver services like IoT applications or end-to-end network slicing. As networks shift to being more software-driven, maintaining and assuring them calls for new skills. On top of this, emerging technologies such as enterprise APIs and agentic AI add layers of complexity but also opportunities.

These technologies enable more granular service differentiation and allow prioritization of business KPIs beyond simple uptime. With the right setup and operational methods, networks can support a wider range of use cases with improved reliability.

How Huawei Supports Intelligent Network Operations

Huawei Intelligent Operations collaborates closely with its network engineers and industry standards organizations to evolve network services and embed best practices. Working with the TM Forum, Huawei integrates customer feedback and its operational vision to develop service-centric performance metrics.

The company invests heavily in AI research focused on telecom-specific digital twins and foundational models. Huawei also applies its extensive telecom network operations experience to refine its own best practices.

Real-World Impact

• Asia-Pacific: Partnered with a carrier to analyze faulty sites using Huawei’s Expected Demand Not Served algorithm. This identified lower-than-usual service usage linked to faults and prioritized fixes, reducing traffic loss by 15%.
• Middle East: Improved FTTx network topology accuracy through delay-tolerant network architecture. This cut invalid work orders by 60% and reduced customer complaints by 10%.
• Europe: Applied generative AI to cut mean time to repair (MTTR) by 25%. The solution involved role-based copilots for field teams and multi-agent collaboration to handle complex scenarios more effectively.

Generative AI in Operations: Applications and Challenges

Generative AI (Gen-AI) has introduced two main operational models: copilots and agents. Copilots assist human staff with tasks like coding, system configuration, and managing service data or trouble tickets. Agents focus on specific operational tasks and can act independently but under human oversight.

Despite the benefits, deploying Gen-AI comes with hurdles:

• Hallucinations: AI can generate incorrect information. Addressing this requires strategies like model training, retrieval-augmented generation, and cross-verification to limit errors.
• Cost: Running Gen-AI at scale can be expensive. Communications service providers (CSPs) need cost-optimization approaches, such as selecting the smallest effective language model and minimizing token usage.
• System Integration: Combining Gen-AI interfaces with digital twins and existing BSS/OSS systems is complex but essential for automation.
• Data Management: Creating digital twins requires breaking down data silos, cleaning and normalizing data, and consolidating various formats into a unified repository.
• Training: Without expert domain training, large language models can’t grasp CSP-specific documentation, tickets, or processes effectively.

Introducing DeepSeek: Efficient Gen-AI for Telecom

In January 2025, Huawei launched DeepSeek, a more efficient Gen-AI model tailored for telecom operations. Research shows DeepSeek excels in domain knowledge application, code generation, data analysis, and intelligent ticket processing—all critical for CSP operations.

Huawei is actively applying these advancements to telecom infrastructure, pushing forward intelligent automation.

Looking Ahead: Huawei’s Vision for ICT Operations

ICT technology remains vital to the digital economy, with new technologies increasingly acting as accelerators. Reliable and intelligent network operations will only grow in importance.

Huawei continues to invest in R&D to support these developments and accelerates the shift to intelligent operations. Collaborating with TM Forum and leading carriers, Huawei recently released the New-Generation Intelligent Operations White Paper 3.0, offering insights into evolving business practices and technology applications.

For professionals interested in advancing their AI skills relevant to operations, exploring resources like Complete AI Training's courses by job role can be a practical step forward.