)
AI is now a natural part of how we work at Nordic Web Team. To make sure this change happens properly, our CEO, Jonas Wärngård, introduced an AI Lead role. AI can indeed create code in a few minutes, suggest fixes, and support development, but some companies might forget that it still cannot understand real-life context, such as deadlines or complex technical decisions.
Our team doesn’t treat AI as an auto-solution. We understand that AI needs clear rules, ownership, and guardrails. This is where our human framework comes in: deciding when to use AI, where to trust it, and how to control it. This framework is led by our DevOps and AI lead, Benjamin Ramnek, whose job is to ensure that AI supports our developers and delivers value to customers without compromising quality or control.
Benjamin’s Philosophy
The best way is to avoid falling into the trap of chasing AI trends and instead, focus on building systems that work for developers and clients.
We don’t chase AI trends, we build systems that actually make sense for our people and clients. AI should expand what we already do well, not replace it; it is about clarity, not chaos. Artificial intelligence becomes meaningful only when guided by human structure, empathy, and experience. Our goal isn’t just to use AI, it’s to integrate it responsibly into how we work and deliver value. – Benjamin Ramnek, DevOps and AI Lead, NWT.
Our goal is to create an environment where using AI feels natural, safe, secure, and helpful. Our team is strong because we balance guardrails, autonomy, structure, and creativity.
Workflow and Development
Before adding AI to any task, we ask: Does this actually improve our workflow? If the answer is yes, then we use AI, but only to make good development practices that simplify tasks. For example:
To help documentation stay up to date.
To make testing become part of the natural flow.
To save time on repetitive tasks.
In a nutshell, AI is here to guide our work, not the other way around. This can include supporting exploration, testing, and writing faster, but the main condition remains that humans are responsible for the accuracy and quality of the final result.
The Framework That Makes AI Safe To Use
Behind the scenes, we have a structure that keeps the use of AI responsible and secure. We focus on:
Billing limits and budget alerts that keep exploration safe.
Access controls that protect sensitive data.
Metrics that measure real performance, focusing on useful gains, not vanity usage.
Quiet discipline that allows creativity to scale.
This turns AI adoption into a structured collaboration between human judgment and machine speed.
What Developers Say About AI
At NWT, AI is constantly evaluated by the people who use it every day: developers. Below are some of what they say about the usage of AI.
Copilot helps with repetitive code. It’s like autocomplete on steroids. — Adrian Grigoruta, Frontend Developer
If you know what ‘good’ looks like, it saves time. If you don’t, it just amplifies mistakes. — Sergiu Moscalu, Full-stack Developer
Sometimes it’s confidently wrong — but still useful. — Alexandru Mazalu, Full-stack Developer
In Magento, it gives me the correct code faster than I could write it myself. For standardised tasks, that’s a big win — I can deliver more in less time. — Radu Istoc, Magento Developer
Having different people with the same concept leads to one conclusion: that AI doesn’t replace craftsmanship. It amplifies it by clearing the repetitive work so humans can focus on the parts that truly require thought.
The goal is to make AI part of everyday development — so it quietly improves results without interrupting the flow. Developers often feel their way is faster, and that’s okay. We want AI to prove its value in real use cases — not be forced on anyone. AI shines in the grey areas — exploring new approaches, improving documentation, expanding test coverage. It’s there to support, not to take over. We’re still exploring what widespread adoption looks like — but we already know where it adds real value.”– Benjamin Ramnek
Rules of Engagement: Where AI Is Welcome (and Where It Isn’t)
AI is powerful, but we always set boundaries:
Security: This is never up for debate. Anything related to security is handled with extra care and strict control.
Critical processes: Humans come first. AI can support, but people make the final calls.
Live debugging: We don’t hand this to AI. The real-time context is too important to risk misunderstanding.
Unclear specification: If something isn’t well-defined, we don’t let AI automate it. We fix the message before we automate the task.
In these areas, humans stay fully in charge. We treat AI like a fast and eager team-mate, who is helpful and always supervised.
The Real Concern
Behind the optimism about AI lies a shared realism on the team that dependence on a single tool is risky.
In the long run, I worry about how companies make themselves dependent on certain AI tools, and suddenly a key feature moves from ‘Enterprise’ to ‘Enterprise Plus’ at triple the cost.– Martin Sternelius, Senior Developer, NWT.
To stay flexible, we build open standards, vendor-agnostic solutions, and systems that continue to work even if tools change. Because we want to own our innovation rather than renting it.
Stay Calm
AI will continue to change the way we work, but we believe in taking a calm and thoughtful approach. Structure first, creativity second, but never chaos!.
AI speeds up the process, but we still choose the direction. Technology becomes powerful when guided by human knowledge and care. That is how we continuously use AI at Nordic Web Team: with confidence, responsibility, and a steady hand.
)
)
)