From Shadow AI to Scale with Meko ElMekawy: How enercity transforms internal AI adoption into a scalable operating model
In this episode, Anne Schlösser speaks with Meko ElMekawy about how Enercity is successfully driving an AI transformation within a regulated environment and what it takes to scale "Shadow AI" into a structured, company-wide operating model.
The conversation centers on Enercity’s "AI Factory" approach, which treats initiatives like a venture portfolio with varying risk levels to drive internal innovation. Meko shares how a baseline survey revealed a massive "Shadow AI" adoption—including the discovery of 400 autonomous agents—which triggered a strategic shift from strict control to transparency and alignment. Using concrete examples like a "Google Maps for utility planning" to optimize infrastructure routes and an "AI Buddy" designed to capture retiring expert knowledge, the dialogue illustrates how AI acts as a key accelerator for the energy transition. Meko explains how framing AI as an operating model and partnering early with the Works Council converts skepticism into high adoption rates, achieving over 90% active usage for tools like Copilot. Finally, he outlines the pragmatic "build vs. buy" decision framework and offers a future vision where AI becomes invisible, critical infrastructure—just like electricity itself.
About the Guest
Meko ElMekawy leads the AI transformation at Enercity, a major German utility provider supplying energy to hundreds of thousands of households. With a clear mission to scale AI as an essential part of the company's infrastructure, he focuses on combining technical innovation with cultural change management to empower a workforce of over 3,000 employees.
Transkript
Meko ElMekawyThat's exactly why we built AI Factory. Not just to drive internal innovation, but to treat AI initiatives as a portfolio with different risk levels, time horizons and paths to scale. Like a venture portfolio. We know that not everything should scale. Some ideas are meant to teach us fast, and some we kill early on purpose.
Anne SchlösserDas ist der Return on Innovation Podcast mit Anne Schlösser und Heike Löffler. Hello and welcome to a new episode of Return on Innovation. Today I'm joined by Meko ElMekawy. He leads the AI transformation at enercity, a regulated utility with more than 3,000 employees, and it supplies energy to hundreds of thousands of households. You might remember enercity's CEO Aurélie Alemany from an earlier episode. She shared enercity's clear focus on renewables, carbon-free heating and smart flexible energy services for cities, businesses and prosumers. And if you know me a little bit better, you'll know that beyond advising clients on AI and growth, I'm also leading our internal AI transformation here at hy. And in this episode, Meko and I reflect on what it takes to bring AI to life in two completely different environments, a regulated utility and a boutique consulting firm. We share lessons, hacks and failures to watch out for on front lines of transformation. Meko, great to have you on the podcast.
Meko ElMekawyThanks, Anne. Excited to be here.
Anne SchlösserMeko, you joined enercity November 2024 with a clear mission from Aurélie. What did the starting point look like and what did she task you with?
Meko ElMekawyExactly. Aurélie, our CEO, had a strong vision. In order to drive the energy transition, we needed AI at scale, and we needed as an essential part of our infrastructure. But as importantly, also to bring people with us. So, back then when I joined enercity with no team, no governance, no infrastructure, and a lot of smart people watching carefully what this would mean for them. And today, we have built an AI Factory, rolled out Copilot with 92% active usage, and run about 20 MVPs and growing of operationalized use cases, all within a very transparent pipeline that the whole company can access.
Anne SchlösserAnd, Meko, when we talk about the AI transformation, people often focus on tools and agents, but often starts much earlier and even without permission. One of your first steps was to take a baseline survey, and this is something we do in our client work as well. Why do you also believe that this is so important?
Meko ElMekawyRight. Before introducing any tools or rules, we ran an anonymous internal AI maturity survey, which is designed to run every six months, to find out where people stood in terms of adoption, personal usage, and also to measure the impact of the tactics we follow from our AI strategy. And the result of the first survey was really eye-opening. Over 50% of our workforce was already using GenAI tools without any approvals or training or governance. And that was the moment we realized, oh, we are not really rolling out AI, we are catching up with the people.
Anne SchlösserYeah, that's quite a surprise, I would say, to find out that so many people are already using it. For us at hy, our impulse came about a year ago when in the partner round we decided that we should obviously also advise our clients on AI transformation, that we are going through it ourselves. So, we actually launched as a first kind of measure an internal AI challenge in order to augment our staff. So, similar, I think, a little bit to your goal. And what we did was that every team member had to build an agent tied to their own work, and they had to make it so that it either accelerates their work or makes it better or, I mean, I guess in the ideal case, both, right? And the goal wasn't just to learn the tools, but it was more to find a way to understand what this technology is changing in organizations like ours, right? And what we learned was really significant. Beyond the fact that the team understands now where agents make sense and where they don't, we also understand much better what we should actually build ourselves and what not. So, that was kind of our learnings. So, and once you understood, Meko, what was already happening inside your company, how did you start driving the adoption further safely and at scale also?
Automatisch erstellt, kleine Fehler sind möglich.