Over the past few months, we’ve been busy rethinking, reshaping, and - to some extent - unlearning how we work. Not because what we had wasn’t good - it was. But because a special tractor, GenAI, showed up in the middle of our farming fields, if we want to stretch an agricultural metaphor of our work as a digital innovation agency.
Even before GenAI, Enrian was never the farm of separate specialised workers each focusing strictly on their own task. Roles blended, people adapted, collaboration preferred to “handovers”. We worked the field together, with our hands and our well known tools.
Then came GenAI. Like a tractor that can till, plant, water, harvest - all at once, and 24/7. Hopefully, it does not fire the farmers, but for sure it requires them to transform.
Today, Product Owner, Business Consultant, and Customer Experience Designer don’t dig. They define the destination. They feed the GPS with insight, constraints, objectives. Solution Architect and Software Engineer make sure the engine runs, the wheels grip the terrain, and the machine doesn’t crash into a wall. UX, Analysts, QA ride shotgun, intercepting nonsense, nudging the outputs, jumping off when needed to fix things the good old way.
“It’s faster, yes. But it’s also at times messier, more intertwined, more demanding. Which is why our Way of Working is in rapid evolution.”
We learnt that this thing needs more than “one driver”. It needs multiple skilled human co-pilots operating in sync, reacting simultaneously. One keeps focus on the business goals, another on the user experience, another checks the respect of good architecture guidelines and so on. And when AI doesn’t understand - a common occurrence - one of the co-pilot needs to jump in and say: "Nope. Let me handle this one."
Stretching another metaphor, we discovered that the engine of this tractor is a genie — one that pops out of a lamp and says, "Your wish is my command". And like in the old tales and jokes, be careful: you get what you asked for — not what you actually needed. GenAI can do marvels, but only with the right context and prompt. Curating them both requires precision & intent. It’s a lot of hard thinking!
We used to work in a very deliberate rhythm with a logical flow: help clients to understand what they need and why → write it down → prototype → iterate → deliver. Now sometimes we start from the other end. We build a working solution first (very fast). Then we circle back, extract from it the documentation. Reverse-engineering structure from output. It feels upside-down, but it’s brilliant. Faster validation and tighter feedback loop to unlock more rapidly business value.
Even fewer handovers in our fields now - Customer Experience Designer, Product Owner, and Software Engineer typing in the same chat window. Even less of “I’ve done my part, here’s the ticket.” and more of “Let’s push this version together, now.” The magic happens when everyone’s sitting virtually on the same bench, hands on the same keyboard.
We have learned the enormous value of having it as a thinking partner. It’s exciting to have automatic generation of code or tests or documentation, but it is just amazing that we shape business ideas, architecture, data model, and processes in conversation with the LLM. We ask, it sketches. We react, it adjusts. It’s not always right, but it’s often helpfully wrong.
It’s not just the way we work that’s changed. Now We Can Grow New Crops. By integrating GenAI in all its forms - chats, agents, canvas, image generation, semantic understanding of unstructured content - we’ve unlocked possibilities we couldn’t touch before. We can build solutions that were not feasible before, simple as that.
Some days, we look at a problem and think, “We would’ve solved this completely differently a year ago. In fact, if we could redo the last ten years of projects, we would.” - and it’s not regret, it’s excitement.
It’s Not All Smooth Sailing.
We’re still learning when to use GenAI to produce prototypes, and when good old design tools are still the best choice. We’re still figuring out the best way to make sure the code GenAI writes is something you’d be happy to debug two months from now; and safe from a whole new type of Cyberattacks. How to have GenAI systematically produce enterprise-grade code - not just prototype-grade, that is easy - or rather to let a GenAI tool to refactor a prototype generated from another GenAI tool.
We will have to master how to retain and grow our talents and what we should look for in the new ones we will hire: if flexibility, multi-skilling and curiosity were once a positive strong “differentiator”, these traits are now a baseline to be effective in our team setup.
We are reshuffling the field activities so that the right people are in the right control seat of the tractor. Some people will need to explore a bit, test new muscles, and find their spike again to fit in the seats. Or walk beside the tractor, doing something the machine can’t (yet).
Like many others, we are undergoing a drastic change in our professional life at Enrian. The implications (of being faster vs slower than others) in capturing a ”3-digits % productivity increase” and an unmeasurable potential in solutions innovation, are disruptive under many dimensions and we take them very seriously.
And yet, despite all this change — and maybe because of it — some fundamentals have only grown more important. They’re not just things we believe. They’re what keep us effective, safe, and human in the loop.
We haven’t changed our DNA. But we’ve upgraded our operating system.
_______________
A Note on How This Article Was Written
This article was written with the help of ChatGPT. Not generated. Written. Co-created, if you want to be generous. If you don’t like it, you would probably not have liked either what I would have written without AI help.
I started with a metaphor - the one with the farmers and the tractor. Then I shared context about how we actually work at Enrian, pulled directly from our Way of Working manual. I also pointed to a doc I’d written years ago with Milan (the original co-founder of Enrian with me) so that the tone and rhythm would match how I naturally write. From there, it became a conversation. I’d draft, challenge, refine. Colleagues jump in and ferociously challenge and comment. AI would ingest it all, suggest, reword, restructure. We brought the experience. It brought the patience and the speed. Together, we moved faster.
This is an example of what GenAI can do: not replace, but collaborate and amplify. And occasionally, surprise you with a turn of phrase you didn’t expect — but are glad to have found. We didn’t always agree, by the way. Sometimes I pushed back on its suggestions. Sometimes it challenged mine. But that’s where the good stuff lives — in the friction, the clarification, the moments of "wait, why that?" That’s where understanding deepens and writing gets sharper. That’s collaboration — just with a different kind of teammate.
