ML Pub Club

Welcome to the ML Pub Club, where machine learning enthusiasts come together to share ideas, discuss projects, and network over drinks. Each meetup features a new speaker presenting on topics of interest to the community.

Meet the Leaders

Matija is an engineer with expertise in ML and computer vision. At Microblink, he led the ML team, contributing to impactful products, and later headed the AI department. He recently founded Eigenpixel, aiming to bridge the gap between machine learning and industry adoption. As a CroAI board member, Matija co-founded the ML Pub Club.

Q: Matija, what inspired you to start the Club, and how has the community evolved since the beginning?

A: “Croatia has a strong foundation in ML, with our engineers and researchers proving themselves globally through successful AI products and top-tier conference publications. We also have a new generation of talented ML engineers who, with proper mentorship, can achieve even more. Building a community from scratch is always challenging, so we chose a relaxed pub setting and launched the ML Pub Club. The key question was whether there was enough interest among Croatian ML experts to sustain a monthly meetup focused on sharing knowledge. After one and a half years, we now know the answer to that question is yes! So far, we've organized 14 ML Pub Club meetups, each attracting 50-60 participants. Considering it was both useful and fun for the majority of the community and us as the organizer, we continued with the program this year with many more meetups lined up!”

CTO at Phytagora

Ivan is a Co-Founder and ML Engineer at Velebit AI, where he has spent the past five years. With a background in computer science and machine learning, Ivan specializes in real-time ML systems and deployment infrastructure. Before joining the ML Pub Club team, he organized the Data Science Croatia meetup group.

Q: What's the biggest misconception about ML, and what do you wish more people understood about it?

A: “One of the biggest misconceptions about ML is that you can simply feed data into an algorithm, and it will learn patterns with minimal effort. In reality, one of the most challenging tasks is cleaning and inspecting the data before using any ML algorithms. It's crucial to ensure that labels are consistent and make sense to humans because if they don't, the model probably won't learn effectively either. Data records can be vast, with numerous attributes. A good data scientist must figure out which features are relevant, how to prepare them, and how to use them effectively to present results. When it comes to modeling with ML algorithms, it takes significant experience to choose the right model and architecture. There are many models, architectures, and approaches, and as models get more complex, they often require a lot of fine-tuning during training.”

Co-Founder and ML Engineer at Velebit AI

Davor is a seasoned business leader with a background in product management. He leads a small team at Velebit AI, delivering AI services across various industries. A serial entrepreneur, Davor has also spent time in academia and the corporate environment.

Q: Davor, what excites you most about the future of ML, and how do you see it impacting everyday life in the next five years?

A: “Product discovery, knowledge sharing, and online search are about to be changed by ML methods. We’ll interact with these processes more naturally and intuitively in the future, which is what excites me most at this moment. Since online search is something we all do daily, both in business and personal life, this shift will impact our everyday routines more than people realize. Future search queries will feel like talking to a well-informed salesperson in the high-end fashion boutique that not only answers our questions but also asks further refining questions and offers recommendations of products we might like to purchase. This change in how we search and navigate digital spaces could radically shorten learning curves, helping us grow and acquire new skills faster, while also making us more efficient in finding what we need in our daily lives.”

CEO at Velebit AI