News

A practical, phased approach from exploration to intelligent automation

At Viatel Technology Group, we made a conscious decision not just to talk about AI, but to lead by example. By positioning ourselves as “Customer Zero,” we committed to adopting, testing, and scaling AI internally before guiding our customers on the same journey.

What followed has been a structured, business-led transformation, focused on delivering real value, underpinned by strong governance and security.

Eilish O’Connor, Director of Strategy & AI CTO at Viatel Technology Group, explains:

“For us, being Customer Zero was about getting hands-on with AI ourselves first. We wanted to understand where it could genuinely add value, where the risks were, and what was needed to roll it out properly across the business. We started with practical use cases, built the right governance and security around it, and learned as we went. That has given us a much stronger foundation to scale AI in a way that is useful, responsible and relevant to how our teams actually work.”

1. Starting with Vision and Purpose

From the outset, our approach to AI was grounded in clear strategic intent. We defined a simple but powerful ambition: to drive internal excellence while establishing external leadership in AI.

Microsoft Copilot was selected as our core platform, with a focus on embedding AI into everyday workflows rather than treating it as a standalone initiative.

Equally important were the principles that guided us:

• A responsible, ethical approach to AI
• A security-first mindset
• A focus on tangible business outcomes, not experimentation for its own sake

Key takeaway: AI must be tied to real business value from day one.

2. Building Governance from the Ground Up

Governance was not an afterthought, it was built in from the beginning.

We established a cross-functional AI Steering Group to oversee strategy, prioritisation, and direction. At the same time, we developed clear policies, guidelines, and usage frameworks, embedding these into both our ISMS and HR structures.

Security and data protection were central to this phase. Leveraging Microsoft Purview, we enhanced our data classification, introduced DSPM (Data Security Posture Management) for AI, and implemented monitoring to identify potential risks, including sensitive data exposure.

Key takeaway: Strong governance and security foundations are critical to sustainable AI adoption.

3. Testing What Works: Pilot and Experimentation

With the right structures in place, we moved into practical experimentation.

Cross-functional working groups were established to identify use cases, map processes, and define measurable outcomes. These were tested through short, focused pilot cycles typically four weeks in duration.

This structured approach ensured that exploration remained grounded in real business scenarios, rather than theory.

Not every idea progressed and that was intentional. Some were paused, others refined, with only the most impactful use cases moving forward.

Key takeaway: Start small, test quickly, and validate value before scaling.

 4. Driving Adoption Through Enablement

Technology alone does not drive transformation, people do.

In parallel with our pilot phase, we invested heavily in enablement. This included structured training programmes covering AI fundamentals, Copilot usage, prompting techniques, and the basics of agents.

We complemented this with workshops, internal resources, and a central AI hub, ensuring employees had access to guidance, use cases, and best practices.

This approach helped establish a shared understanding across the organisation covering not just capability, but also risk, responsibility, and data awareness.

Key takeaway: Adoption requires the right balance of skills, confidence, and guardrails.

5. Scaling What Works

As confidence grew, we began scaling successful use cases into day-to-day operations.

These ranged across the organisation from document creation and marketing content to data analysis, tender support, and finance processes.

The results were clear:

• Significant time savings (in some cases reducing days to hours)
• Greater consistency and quality of output
• Improved efficiency across key business functions

At this stage, AI became embedded within our broader company strategy, not treated as a standalone initiative.

Key takeaway: Real value is realised when AI becomes part of everyday workflows.

6. People & Culture: A Human-First Approach to AI

As conversations around AI increasingly focus on job disruption, our approach at Viatel has been clear from the outset: AI should empower people, not replace them.

We’ve taken a human-first approach, ensuring that adoption is not just about efficiency, but about enabling our teams to focus on more meaningful, higher-value work. By automating repetitive tasks and enhancing day-to-day workflows, AI is freeing up time for creativity, problem-solving, and strategic thinking.

But technology alone doesn’t drive this shift, leadership and culture do.

To support our teams through this transition, we’ve focused on open communication, continuous learning, and creating space for people to adapt at their own pace. A useful framework that has guided our thinking is the “5Cs” model, highlighted by Ryan Roslansky and Aneesh Raman, which focuses on building the human edge in an AI-powered workplace:

These principles have helped guide meaningful conversations across the organisation, ensuring we address both the opportunities and concerns that come with AI adoption.

Ultimately, our goal is not just to introduce new technology, but to support our people in evolving their roles shifting focus from manual tasks to more impactful, rewarding work.

Key takeaway: Successful AI adoption is as much about people and culture as it is about technology.

7. Moving to a Mature Operating Model

As our approach evolved, we shifted from broad experimentation to more focused, project-based delivery.

We introduced formal prioritisation frameworks, ROI tracking, and regular reporting at board level ensuring visibility and alignment at the highest levels of the organisation.

Training also became more targeted, with increased focus on advanced capabilities such as Copilot and agents, while reducing unstructured exploration.

Key takeaway: Scaling AI requires discipline, structure, and alignment with business priorities.

8. Entering the Next Phase: Automation and Agents

Today, our focus has shifted towards automation and AI-driven execution.

We are actively developing Copilot-powered agents to support:

• RFP and tender processes
• Workflow orchestration
• Knowledge discovery and retrieval

This marks a significant shift from AI supporting individuals to AI actively executing tasks and processes.

Key takeaway: The future of AI lies in intelligent automation, not just assistance.

 9. Exploring Beyond the Core Platform

While Microsoft Copilot remains our foundation, we continue to explore complementary tools and platforms where appropriate.

This includes testing emerging AI technologies and domain-specific applications across areas such as security, networking, and data.

Key takeaway: A strong core platform enables innovation at the edges.

10. What Comes Next

Our journey is ongoing.

Next steps include:

• Expanding high-value use cases across the business
• Scaling agent deployment at process level
• Continuing to strengthen data readiness and governance
• Enhancing ROI tracking and transparency
• Driving engagement through ongoing initiatives such as workshops and hackathons

Our long-term vision is clear: to become a fully AI-enabled organisation and a trusted partner supporting customers on the same path.