In a time when the digital sphere is flooded with AI content and there is an overproduction of texts, images and videos, which have almost no content value for users, it is a challenge to write a blog post that will stand out from all that noise. Thinking about this topic, a scene from one of my favourite series “Star Trek” comes to mind, where the crew of the spaceship Enterprise is attacked by superior aliens, who can transform into any member of the crew. No scanner can recognise whether it is a real person or an alien disguised as that person.
In one scene, Captain Picard and First Officer Riker meet. To find out if they are the real ones, they share each other a memory from their youth that only they know. Thus, this creative solution ultimately proved to be decisive in dealing with the aliens.
The New Chapter of Tivius: Complete Digital Solutions for the Modern Era
In this context, a few months ago, while talking to several clients for whom we were building websites, I asked them what real problems they had that could be automated. For several months before that, I had been reviewing, researching, and learning about the possibilities of AI agents and automation. After the conversations, I realized that there were broad areas in the operations of companies that could realistically be optimized and automated. And that was not to replace employees, but to optimise processes that would allow employees to focus on more productive tasks, and companies to become much more efficient and profitable.
So, after those conversations, we focused on learning, researching, and building systems based on artificial intelligence that would be able to solve the listed problems of companies, in terms of manual processes that could be automated. After a few months, we managed to create a structure, knowledge, and functional models that solved those problems. This is where the idea was born that this could be, not just a new Tivius service, but a completely new agency, which we called Tivius AI Agency.
The Technical Architecture of Efficiency
But let’s get back to the problems and how we came up with solutions. The clients we spoke with had heard about the possibilities of artificial intelligence, specifically about popular LLM tools like ChatGPT, Gemini, Claude, etc. Many of them used them in their work. More than a few clients had also heard about AI agents. But in reality, none of them had the technical expertise, let alone the time to train, build, and integrate an AI agent into their work. We saw the opportunity here.
I had been studying n8n for several months, which was installed on my vps (virtual private server) on Hostinger. Driven by the ideas, I built a structure of several, at first automations, and later agents. Then, through API, I connected tools like G Drive, Gmail, Google Sheets, etc. In the training structure, I used a vector database, and as an activator (trigger), I used web-hooks. I did several tests and presented the results to the clients. The real outcome was more than excellent. The agents and automations themselves performed tasks in seconds that previously took customers hours to complete manually.
The bridge between technology and business: Why are new AI agencies crucial?
And this is precisely what will gain momentum all over the world. The rise of agencies, which act as intermediaries between new technologies related to artificial intelligence and companies. And above all, companies that have the challenge of independently integrating these technologies into their processes.
According to official Eurostat data, only 20% of enterprises in the EU used at least one AI technology in 2025. In Macedonia, this figure is strikingly lower — only about 8.5% of businesses use artificial intelligence in their operations.
In our estimates, the next few years, the demands for these services will have explosive growth. And not only for companies, but also for all types of organisations and institutions.
Risk Management: How to Automate Without Losing Control
However, all this has another dimension that must not be neglected. In a recent study, titled “Agents of Chaos,” scientists and researchers from several renowned universities (including Harvard, Stanford, MIT, and others) tested autonomous AI agents. They haven analysed several scenarios, and the agents exceeded the limits set for them. In addition, they committed illegal actions such as hacking, leaking personal data, unauthorised downloading, etc.
This is an additional reason why a kind of intermediaries and controllers are needed. They will set restrictive limits and supervise these AI agents. As great as the benefits of AI agents are, the greater the risk of their improper and unrestricted deployment.
Conclusion: Our mission towards increased productivity
Finally, that is the mission of Tivius in the coming period. To enable companies and organisations to integrate AI agents and automations into their operations. And that, in a safe and controlled manner. Thus, they will significantly optimise processes, but also the entire society will increase productivity several times.
Author: Blagoj Gjelevski