„ON THE LADDER OF SUCCESS“ WITH PROF. DR. GJORGJI MADZAROV: AS A STUDENT, MY PASSION WAS SPORTS, AND NOW IT IS ARTIFICIAL INTELLIGENCE
“Passion is the man motive, he does what he wants to do, no matter what. “As a student, my passion was sports, and now it is artificial intelligence,” says about his life path Gjorgji Madzarov, a FINKI professor and researcher in the field of artificial intelligence.
Professor Madzarov works in the field of machine learning and software engineering and is one of the founders of the startup company “Elevate Global”, which works predictive modeling for the energy sector needs. Madzarov never thought he would have his own company to promote on the world market, especially not while throwing a basketball at 40 °C in his hometown Veles.
Photo: Private archive
– My story started from shooting the basketball, where the basket was placed on the pole on the City bazaar.
My idols then were Michael Jordan and the Chicago Bulls, and sport was my biggest passion. The first example of what I should be like was left to me by my gymnastics coach – Trajche Necinov. While the other teachers hung out with each other, Trajce sat with us children, and told us stories about his life journeys. That, left a strong impression on me. Building an attitude of respect for others, the elderly and the more experienced. You can learn from everyone, and even from the one who is laying you can learn something, you can learn how to lay down – Professor Madzarov remembered with a smile.
Through sports he builds the character of patience and good judgment, needed in computer science
He is one of the few scientists who built the way of life and success on sports and always emphasizes that thanks to him, he built a character of patience and good judgment, later needed in computer science.
– I built my spirit with the sport lifestyle, with the rhythm of going to bed and getting up while playing volleyball or while training as a lifeguard. We knew how to compete in every game, from street to school. Today we miss that the most. I was always into sports, I just loved it and I was never too attached to study. I was studying as much as I needed to be good, to be great, but not to be the best or the most outstanding. Sport is a healthy spirit, perception and correct judgment in difficult moments. When it’s hard for me, I draw most strength from sports. It is very difficult for a sport person who has gone through all the stages in sports, to make him fall mentally – says Madzarov about himself.
Competition in knowledge
His interest in computer science appeared in high school, in the mathematics class where everyone competed with knowledge.
– A large part of that group went to Faculty of Electrical Engineering because we thought that there was the future. When we saw what computer science could achieve, which was in its infancy here, we fell in love with it.
Discipline and brain gymnastics
When he started studying at the Faculty of Electrical Engineering and Information Technologies, he realized that he had to stop training at the volleyball club “Vardar” in order to dedicate himself to the exercises because then, a lot of time was spent in the laboratory.
– At that time there were no resources on the Internet, our main source was the professor’s books or scripts. I was disciplined just like in sports, except that this was brain gymnastics. I want to learn new things, so I did not miss a teaching class. There are no limits for man. I finished the first year with an average of ten, unexpected for me and my parents, who sent me to study with a “bequest” to finish it faster and not to delay because the finances are limited – says the professor.
In Slovenia I saw what a research institution looks like
After graduating, Gjorgji Madzarov got a job at Netcetera. He stayed there for only three months, after which he left to be a university demonstrator. The two years of master studies followed, after which he received an offer from Slovenia for doctoral studies, but he decided not to leave and write a doctorate here in Macedonia. During his doctoral studies, Madzarov was on a one-year stay in Slovenia with Professor Saso Dzerovski.
– In Slovenia I saw what a research institution looks like, which we do not have here and will not have for many years to come. I came back here, got my PhD and started a career at college, as an assistant professor.
How the AI company was formed
Professor Madzarov never thought he would have his own company to promote on the world market for artificial intelligence. Today it is a reality in the phase of aggressive marketing.
Netflix co-founder Mark Randolph once said, “Nobody knows what he’s doing.” It ‘s like running trough water and you do not know where the river will throw you. The idea came from a colleague of mine, Goran Stojanovski, and it happened somewhere at the end of 2018. He was an assistant professor at the Faculty of Electrical Engineering, and in the meantime I switched to FINKI and we wanted to see each other over a coffee chat. He told me, “I want to start a company because I see artificial intelligence is the future.” We made combinations, started some research with the help of fellow students, started working, got an idea and presented it in several places. We are now in the phase of aggressive product marketing and we are literally waiting to see if we can rise to the level of becoming a global company. We are currently present locally and regionally, but we want to enter the world market relatively quickly.
Madzarov and Stojanovski presented the idea in Berlin at the end of 2019 and won first place. Then they were supposed to go to the United States, but the pandemic happened and they started working here, and they managed to get funding from the Innovation and Investment Fund from the UKIM accelerator.
Software – a kind of energy Wall Street
– “Elevate Global” manufactures products that include the need for artificial intelligence and machine learning to predict time series of energy data. This means that our task is to play the role of a modern granny Vangja and to enable the optimization of the amount of energy produced in relation to the demand. This is not about predicting the amount of energy produced by thermal power plants or hydropower plants, for which their production can be planned very precisely. It is about energy produced from renewable sources and wind farms. With these sources, production depends on the weather and it is very unpredictable how much energy you will produce. The goal is to use the energy produced from renewable sources and only what we lack to supplement the traditional energy sources. And the more accurately we predict, the more accurately we will be able to plan in the coming days how much energy we will need from traditional sources to make up for that shortfall. If the prediction is inaccurate, we will have a greater shortage or surplus of energy, energy that we will have to provide (buy) or energy that we will have to store. It is a waste of energy that additionally contributes to increasing the carbon footprint – explains Madzarov.
The software, which is a kind of energy Wall Street because it can predict and optimize energy consumption, processes a huge amount of data in the form of time series. The Elevate platform works with the help of domain knowledge that has been introduced by the team thanks to the knowledge of how the energy networks work.
– If we reproduce the same model for financial data, it will not work because the model has no domain knowledge of financial data. The question is why not do it for financial data? They depend on many factors. A statement from an influential person can drastically affect the market. Financial markets are highly volatile, predictable, but quality is low and small turbulences can bring losses. In the energy world, everything is due to the laws of physics, not someone’s statement. Additionally, our platform was used to process 350,000 protocol messages per second over a telephone network to monitor the quality of VoIP call service. This was a project in Germany where we monitored and analyzed the traffic within a whole network – added Madzarov.
In a pandemic, students are lack of competition
Although he is an expert in computer science, Dr. Gjorgji Madzarov believes that the virtual instruction imposed by the Covid-19 pandemic has not contributed to improving the education system and that nothing can replace direct contact and interaction in the classroom.
– You teach on a computer, not to students. You only see the number of students, 20, 30, 50 or 100 students who do not know what they are doing at the moment. Often they just typeing without communicating. They just watch the presentation, listen to the lecture and that’s it, unfortunately. When they come to take the exam it is even harder, small screens, internet connection drops, etc. Additionally, the discrepancy among students is increasing, 3-5% of them are superior, there are no limits for them. But the core group that should be the core is getting lost, the gap is getting bigger. They lack a competition, a challenge – he explained.
In the vineyard with his father, in the Veles bazaar and the promenade …
In his free time, Gjorgji often rides a bicycle with his daughter and it is his only sport activity after his injuries prevented him from playing volleyball and basketball. Gjorgji often comes to his hometown Veles to hang out with his parents and friends. He wants to work in the vineyard with his father, to walk through the Veles bazaar and the promenade, and he does not miss passing by the basketball court from his childhood, which now, unfortunately, shines empty.
Author: Orce Kostov
