FCSE · Ss. Cyril and Methodius University · Skopje
I build machine learning systems that work on real data, from satellite imagery and medical scans to search ranking and on-chain markets.
I am an Associate Professor at the Faculty of Computer Science and Engineering in Skopje, where I lead research and industry projects, lecture, and mentor students. My research sits where machine learning meets large, messy, real-world data: Earth observation and remote sensing, medical imaging and retrieval, and fairness in search.
Alongside academia I have spent more than fifteen years building production systems. At Permian Labs I single-handedly designed and built the NFT collateral analytics that backed over $60M in loan volume. I ran engineering for a retailer-to-carrier shipping platform in Amsterdam, led the recommendation and search team at Infinite Analytics, and founded two companies of my own.
I like problems that need both halves: a model that is only useful once it is wrapped in a reliable API, a pipeline, and a dashboard someone actually opens.
Citation metrics from Google Scholar, November 2026.
A toolbox of state-of-the-art machine learning methods for exploratory and predictive analysis of satellite imagery, with a repository of AI-ready Earth observation datasets. It covers land use and land cover classification, crop type prediction, semantic segmentation and more, and underpins an open benchmark arena for EO image classification.
A system that gathers market data from many sources, models the collateral value of NFTs and serves the metrics over REST APIs, plus dashboards for the protocol and the wider NFT lending market. It was used to generate over $60M in loan volume.
Led development of a multilingual model that detects forest-fire content in Twitter data, integrated into SMDRM, a larger system for crisis management.
Open-source tools for fair ranking: FA*IR post-processing and DELTR in-processing, shipped as an Elasticsearch plugin and Java and Python libraries. Presented at The Web Conference 2020.
Led the team that designed and ran a platform where retailers enter shipments and carriers are assigned automatically by business rules, with every party notified along the way.
A camera network that measures average vehicle speed and detects violations, and a multimodal biometric authentication system combining face, fingerprint, retina and voice.
Earth observation
Agriculture
Medical imaging
Fair search
Full and up-to-date list on Google Scholar.
Leading and participating in research and industry projects, lecturing and mentoring students. Assistant Professor 2017–2022, Research and Teaching Assistant 2011–2017.
Sole engineer: designed and built the NFT collateral valuation system behind MetaStreet's lending protocol, which served $60M+ in loans.
Multilingual social media model for forest-fire crisis management.
Led design, build and operation of the retailer–carrier shipping platform.
Led development of the FairSearch fair-ranking libraries and Elasticsearch plugins.
Led the Macedonia team behind the recommendation and search APIs, ETL synchronization, traffic reporting and monitoring, working with clients and teams across time zones.
Data science and engineering consultancy.
Cross-exchange banner network used by more than 200 websites.
Early teaching and research in computer vision, and a network traffic reporting system.
For research collaborations, student supervision, or consulting on data and AI systems.