A telecoms manual in the back pocket of every network engineer: Why the Wray Castle knowledgebase is being uploaded into Metricell’s AI-native platform
This article was originally published by The Mobile Network.
AI agents without context are like network engineers without training. So, why not give the agents access to the same resources used to train the engineers?
Most telecoms professionals have graduated from one or more of Wray Castle’s courses during their career. Operating since 1958, the training provider has upskilled over 300,000 professionals across 85 countries and offers courses in everything from fundamental networking to cutting-edge 5G, AI and mission-critical communications.
Now, Wray Castle is partnering with Metricell to turn these resources into an intelligence layer for Metricell’s agentic AI platform. Agents will have access to the relevant content library, including over 1,720 hours of training content across 20,640 topics, to provide the technical context needed to interpret network data.
A Shared Telecoms Manual
Operators are not short of network data, but interpreting it requires extensive knowledge of network behaviour, performance thresholds, common fault patterns and proven troubleshooting approaches. Without the appropriate training, humans struggle to understand what the data means and AI agents produce conclusions that sound confident and fluent, yet ultimately incorrect.
The Wray Castle and Metricell partnership will enable AI agents to draw on a resource created for experts, by experts. Agents are provided with context in the same way human engineers are provided with training, ensuring their responses are grounded in established telecoms practice.
We are not aiming to replace experienced network engineers with AI agents. The best telecoms engineers have contextual knowledge of the sites they manage and many years of hands-on experience, so they will continue to play an essential role in the management of the network. Instead, our partnership will support engineers by providing direct access to the Wray Castle resources within the Metricell platform.
Even the most experienced engineers encounter unfamiliar technologies or exceptional use cases that require additional guidance. Our agents present the evidence behind their conclusions, linking engineers directly to the relevant training video or section of a
document used during the investigation. This gives engineers the confidence to verify the agent’s reasoning and deepen their understanding of the topic before deciding what action to take next. The partnership essentially places the entire telecoms manual in the engineer’s back pocket and uses AI to augment human expertise, rather than replace it.
The Intelligence Layer
AI agents in the Metricell platform have access to an intelligence layer containing the Wray Castle training resources, operator policies, technical documentation, user guides and other approved materials. Agents draw on this knowledge at multiple stages of the workflow:
1. Understand: The agent interprets the user's request, loading relevant agent skills, operator policies, technical documentation and Wray Castle resources to understand the task.
2. Investigate: The agent queries the digital twin of the network, retrieving relevant information from network systems, operational records and other connected data sources to investigate the issue.
3. Validate: The agent runs predictions using the digital twin to test potential explanations, comparing the results against relevant technical knowledge and established telecoms practice.
4. Present or act: The agent returns an auditable response, presenting the findings alongside supporting network evidence and references to the learning resources used. This allows engineers to review the reasoning and, where approved, initiate the next workflow.
Supporting Your Best Network Engineers
Consider an engineer investigating the root cause of a network issue. They need to pull together evidence before deciding on a recommended course of action. Depending on the complexity of the issue, gathering this information could take several hours or more than a day.
Instead, an AI agent in the Metricell platform brings together these sources within minutes, identifies probable causes and presents the evidence through data tables, charts and
written explanations which can be exported as an engineering report. Relevant Wray Castle training materials are included in this analysis, providing context on the specific network principle or troubleshooting approach.
Rather than searching across multiple systems or training repositories, the engineer would receive the agent’s findings alongside the supporting operational evidence and relevant technical knowledge. Direct links to Wray Castle documentation allow the engineer to review the technical principles behind the analysis or explore the topic in greater detail.
AI to Improve Human Expertise
The Wray Castle and Metricell partnership creates a shared foundation of knowledge between humans and agents, presenting a unique proposition to the telco market. The project also shifts the burden of training every agent or building a telco-specific knowledge source away from operators’ internal teams. As the industry evolves, new training materials can be incorporated into the knowledge layer to keep it aligned with current technologies.
Together, we are keeping people at the centre of the decision-making process. AI agents surface the relevant operational evidence and technical knowledge, while engineers remain responsible for interpreting it and determining the appropriate response.
Want to learn more? Contact marketing@metricell.com to connect with our team.
.png)
.png)
