Luka Lučić and Damjan Karadžić are two young people from Montenegro brought together by an interest in technology and in solving practical problems. Luka is studying mechatronics at the Faculty of Mechanical Engineering (UCG), having previously completed secondary education in web and mobile development, which shaped his preference for building systems from initial concept through to finished product. Damjan is a secondary school student on the mathematics track, with a strong analytical and algorithmic background that comes through most clearly in data analysis and model tuning. Firefly is a project they build together, end to end, as they put it: “We are building Firefly together, end-to-end, and we are ready to learn as we go.”
Firefly is an AI system for early detection of wildfires, risk assessment and prediction of fire spread. Drones fitted with cameras fly over at-risk areas, and computer vision analyses the footage in real time to identify smoke and fire at an early stage, when intervention is cheapest and most effective. Alongside detection, the team is also working on prediction: using data on terrain, vegetation and conditions to assess where the risk of a fire starting is highest and how it might spread. The problem the system addresses, in their own words: “fires are noticed too late, and the region is increasingly vulnerable to seasonal disasters.”
At the Adria Future Summit, the team presented Firefly and placed second in the hackathon. For them, this was the first confirmation that the idea holds up outside their own team, and encouragement to keep developing it further: “For us, that was the first confirmation that the idea makes sense outside our own team, and encouragement to keep developing it more seriously. It also showed us how valuable it is to stand in front of people, get feedback and defend your solution in person.”
Three months on, the focus has moved from pitching the idea to building it out. The team is refining the model and preparing an MVP for pilot testing with real drones and locations, alongside development of the website and brand at fireflyapp.me. In parallel, they are applying to programmes and competitions that could accelerate the work.
Asked about the state of the regional innovation scene, they see clear potential alongside clear gaps: “The region has strong engineering and software talent, but it lacks capital, mentorship and connections to the market. The opportunity is large precisely because problems like fires, climate and public services are concrete and local: solutions can be tested here and scaled more widely. We need more events that connect people with resources.”
On what the next generation needs most from events like this, they point to practical access over theory: “Less theory, more access to real mentors and investors, and the chance to test an idea on a real problem. The greatest value is networking and the feeling that serious work is being done, not just simulated.”
The connections made at the Summit have held up over the following months. The team stayed in touch with other participants and members of the tech community they met there: “These connections are valuable to us because they open doors to exchanging experience, advice and possible future cooperation, and what matters most is knowing that we are not alone in this.”
Looking ahead to how ideas like Firefly might develop over the next year, and to where AI is heading more broadly, the team expects a shift towards systems that act rather than simply assist: “We think the next wave of ideas will be much more agentic: fewer static tools, more systems that make decisions and carry out tasks on their own. Access to powerful models will become cheaper, so the focus will shift from whether something is technically possible to whether it solves a real problem and is sustainable. We expect more vertical, narrowly specialised solutions.”
