About us
Why we started uvedi.ai
It started with one frustration from the field. Visiting plants and construction sites across Croatia and the region, we saw the same pattern again and again: a buyer looks for a local company, fails to find it fast enough, and the job ends up with a large foreign firm doing exactly the same thing — not because it is better, but because it is more visible. What hit us hardest was how often we watched excellent local companies lose work not on quality, but because the market simply never found them in time.
Our mission
To raise the level of AI literacy in small and mid-sized companies and bring them into the AI era without losing what makes them recognisable: their craft, their people and their way of working.
A tool on its own changes nothing — the change comes from an employee who understands what the tool does, where it fails and when it must not be trusted. That is why the knowledge of the people in the company comes first and the technology second. Technology in a manufacturing company should never be a reason to change what the company already does well. Our job is to give back the time that disappears today into quotes, documentation and explaining the same things over and over — and to get what that company knows in front of the customer before somebody else does.
Customers ask AI first, not you
People have less and less patience, and the old way of doing business through contacts and word of mouth is slowly disappearing. When somebody looks for a tradesman or a supplier today, they no longer ring around — they google and ask AI tools. The younger generations now making the decisions even look for solutions on TikTok. If you are not there, to them you do not exist.
The winner is not always the one with the best product, but the one who knows how to show it best.
Experience from Denmark, brought home
We built our business and engineering experience in Denmark, working alongside companies such as Maersk, Novo Nordisk, Bosch and Grundfos. Working on technologies like Digital Twin systems in VR, we realised the rules for success are the same everywhere, whether it is a global corporation or a top team of electricians on site: quality has to be visible. We decided to bring that knowledge home. We are not a classic marketing agency, because we understand both worlds: we can read a technical specification and grasp the complexity of your work, and translate that quality into a message that earns attention in a few seconds.
Filip Kovac
Solutions Engineer
M.Sc. Technology-Based Business Development, Aarhus University · B.Eng. Global Business Engineering, VIA University College.
LinkedIn profileWe have worked with generative AI since 2019
Our first projects with language models date back to the GPT-2 era, years before ChatGPT became an everyday tool. The work was research and industry at the same time: VR avatars with automated dialogue, real-time speech synthesis and Digital Twin systems for production environments, together with Danish industrial partners.
For a client that experience means one practical thing: we can tell what the technology can genuinely do from what simply sells well. We went through every stage — from models that could barely form a sentence to systems that handle enquiries on their own — so we judge the business case by how the technology behaved in production, not by the trend.
The difference is how long we use a technology before we recommend it. When AI becomes a trend, everybody suddenly offers it; we worked with these tools while they were still unreliable, so we know where they break. A prototype that works in a demo and a system that survives daily work in a company are two different things — we deliver the second.
In the first meeting we say what AI cannot do in your case, or is not worth doing, before we propose a solution.
We do not introduce technology into a process nobody has mapped. First the workflow, then the decision on where automation actually pays off.
Within the agreed scope we hand over the solution documented, with access credentials and an explanation of how it works, so you can maintain it without us.
Research
AI in engineer-to-order engineering
The research was carried out at Aarhus University (Technology-Based Business Development, 2025) in cooperation with the Danish manufacturer C.C. Jensen A/S, a world leader in offline oil filtration for the marine, industrial and energy sectors. The research question: how to use artificial intelligence to reduce manual engineering work in engineer-to-order shipbuilding projects while keeping flexibility and compliance with shipyard requirements.
- Full process mapping (BPMN) from budget quotation and the formal tender phase through to contract, purchase order and factory acceptance test.
- Semi-structured interviews, questionnaires and document analysis with key stakeholders in engineering, sales and project management.
- Building and evaluating an AI prototype following the Design Science Research methodology, focused on the limits of language models and accuracy control.
The same logic drives how we work with clients: first map the process and the bottlenecks, then introduce AI only where it removes measurable manual work.
First work with language models in the GPT-2 era, in research and industrial contexts.
VR avatars with AI automation and Digital Twin systems with Danish industrial partners.
Business applications of generative AI: enquiry handling, documentation and multilingual content.
uvedi.ai — AI visibility and process automation for manufacturing and service companies in the region.
We speak two languages
A marketing agency understands the customer but not your product. An engineer understands the product but describes it in language the customer never finishes reading. We translate between those two worlds — and that is the only reason our copy passes with both your process engineer and the buyer.
Tolerance ±0.05 mm, surface roughness Ra 0.8.
The part fits without rework, assembly never stops on the line.
We work to EN 1090-2, EXC3.
The structure passes inspection and handover first time, with no follow-up paperwork.
We have our own tool shop and CAM programming.
A drawing change does not push the delivery date out by a week.
We do not invent a message. We translate what you already do well into a sentence the customer — and the AI guiding them — understands immediately.
We do the same translation towards machines: technical data, certificates and tolerances written in machine-readable form are the reason an AI search engine can name you as a supplier at all.
What this means for you
We have seen what works at the highest global level, and we know how the ground breathes here in the region. We build your complete digital and physical presence turnkey, bilingually and strategically, for domestic and foreign markets.
You keep doing what you are best at; our job is to make sure the whole world finds out.
LinkedIn ↗Who we work for
Manufacturing and service companies for which visibility matters as much as the product.
How we work
Concrete
Numbers and facts instead of adjectives. Before/after proof for every item.
Fast
First analysis with no waiting. A website in weeks, not months.
Straight
Prices are final. We never sell you anything you do not need.