Building the AI-Powered Telco: How GenAI is Driving IoT Growth

A Challenge Bigger Than Traffic

In the ever-evolving telecoms industry, IoT isn’t a product. It’s a concept. That’s always been the challenge. Businesses know they need better connectivity, more innovative infrastructure, and seamless automation, but they don’t always realise that IoT is the answer.

That’s where the real problem lies. How do you drive IoT adoption when customers don’t even know they should be looking for it?

🤖 AI generated image using DALL-E | an application designed to grow traffic

I saw an opportunity to shift the conversation. What if we let AI do the work rather than relying on traditional marketing and sales? What if we surfaced relevant IoT solutions when a business realised it had a problem without forcing it through layers of research and technical education?

That idea set everything in motion.

Bringing AI to the Forefront

With that vision in mind, I started laying the groundwork. This wasn’t just about AI. It was about bringing the right people, tools, and strategies together to make it work. I kicked off the conversations, negotiated partnerships, and set the direction for what this needed to become. The entire technical feasibility, business alignment, and strategy are combined under one guiding principle: build something tangible that could move beyond an experiment and into production.

Two key solutions emerged: an AI-generated blog that could create engaging, relevant IoT content to drive traffic and conversion and an AI-powered consultation tool that lets customers describe a challenge in their own words and instantly receive a tailored BT solution. The pieces were coming together. Now, it was time to prove they could work.

AI-Powered Content that Drives Traffic

The first breakthrough was automating content generation at scale. EPAM’s InfoNGen AI platform was the foundation, allowing us to scan the web for trending IoT topics, rewrite and refine that content using GenAI, and structure it into blog articles that actually provided value.

The key was making it seamless. Each piece of content naturally integrates calls to action, guiding readers towards self-service ordering journeys without forcing the sale. If BT couldn’t offer a direct solution, the system provided fallback recommendations, capturing potential leads even when BT wasn’t the answer.

This wasn’t just content marketing. AI-driven, targeted lead generation surfaced IoT solutions exactly when customers needed them.

AI-Powered Problem-Solving

The second challenge was even more prominent. What if we could remove the need for customers to understand IoT altogether?

Instead of making them browse endless lists of solutions, we built a tool that lets them describe their business challenge in their own words and let AI translate that into a solution. A business enters its name or website, and AI analyses industry trends, common challenges, and open data sources to build a profile of what it might need. A simple Q&A refines the problem further, and AI matches it to a BT IoT solution—turning a vague issue into a concrete action plan.

Now, instead of trying to explain what IoT is and why it matters, we could meet businesses exactly where they were. The conversation shifted from selling technology to solving real, tangible problems.

The Moment It Became Real

When we first shared the proof of concept within BT, the reaction was immediate. This wasn’t just a tool for IoT. It was a model that could reshape how AI is used across the business. People saw the potential: an entirely new way to drive traffic, a scalable system for AI-led problem-solving, and a faster, more intuitive sales journey. Most importantly, it worked.

Within weeks, fundamental production ideas started emerging. What started as a vision quickly became tangible and could make a real impact.

More Than Just an Experiment

This wasn’t just about building an AI solution. It was about proving what’s possible when you step beyond traditional thinking. It meant navigating the technical landscape to ensure feasibility, aligning stakeholders across the business to secure buy-in, negotiating partnerships to bring the right expertise onboard, and driving the vision from concept to execution. The only thing I didn’t do was write the code. But everything else—from the idea to the execution—was built from the ground up.

And this is only the beginning.

How It Works Under the Hood

This was a fusion of AI-driven content generation, real-time business analysis, and automated lead qualification. The AI-generated blog used InfoNGen AI to scan global content sources, filter by relevance, and generate articles that felt natural and informative. Transformer-based GenAI models rewrote and structured the content dynamically, ensuring every piece was aligned with trending conversations and surfaced the right call to action at the right moment.

The system leveraged a mix of natural language processing and data analysis for the AI-powered IoT consultation tool. AI scraped publicly available business information, profiling companies based on industry, activity, and potential challenges. Machine learning models identified common pain points and mapped them to BT’s portfolio, offering businesses a personalised solution without needing them to self-diagnose their technical requirements.

This entire framework was built with a scalable, API-driven architecture, enabling integration directly into BT’s existing digital ecosystem. That meant real-time updates, seamless content deployment, and AI-powered customer interactions that could evolve and improve with every new engagement.

This wasn’t just a proof of concept—it was a real, working system that changed how we think about AI in telecoms. And it’s only just getting started.

Thomas Archer