It’s been an exciting year for our Ljubljana team

A lot has been happening in Ljubljana since January. Home to our engineering teams, we have grown by 40% and we're still hiring.

With the team expanding, we outgrew our original space so in February we moved into a new office. The new office also gives us space to host team events and networking activities. 

The Hard Part of AI

Last week, we co-hosted the opening event of the session The Hard Part of AI with SLOAI, the leading AI community and professional network in Slovenia. As a new member of this organization, we were excited to kick off the season with a discussion on best practices and learnings from working with AI. 

As all developers and engineers have experienced firsthand how quickly AI can generate code, create demos, and provide surface-level knowledge. But the harder part is understanding, validating, and trusting what AI creates. 

Three of our team members shared their perspectives on this topic

Tadej Borovšak, Director of Engineering, discussed cognitive biases in the era of AI. When an AI tool provides a quick answer to a complicated question or code for a problem you're solving, it's easy to take it at face value without fully understanding the logic behind it. 

So Tadej's advice was simple: take a step back and make sure you can explain the logic behind the results. It’s okay to acknowledge that your understanding of the problem isn't as deep as you thought. When that happens, ask the AI clarifying questions or even discuss the thinking with a team member. The result will be better decision making and the confidence that your work and the logic behind them are sound. 

Nejc Potrebuješ, Senior Software Engineer, explored the real cost of coding faster. AI has made writing code faster than ever, but has also introduced additional complexity. Now the engineering bottleneck has shifted from creation to comprehension, quality, security, and maintenance. Nejc posed the questions: what does it actually cost to code faster? Can we trust the code generated without needing to validate it? Or will we need more AI tools just to understand, review, test, and maintain the code AI creates? 

We may not have answers to these questions but they are worth thinking about and ones we will need to start addressing sooner than later. 

Matej Drev, Chief Operating Officer, shared lessons learned from moving beyond AI demos to solutions that users can trust, use, and will pay for. Customers find value in solutions that address their most critical problems and at the same time can be easily incorporated into the workflows they already use today. Based on his 7 years of deploying AI, he still believes in these core principles: 

1. Start with the challenge
2. Domain expertise matters
3. Success is measured by how much of the problem AI solves

Matej reminded the audience that you don’t need perfect data to demonstrate value as long as you're earning the user's trust along the way. At the end of the day, we know AI will continue to evolve and impact the way we work but delivering real value is what separates a pet project from a paid AI solution.

It was a great event! Special thanks to SLOAI, the speakers, and Zala Hovnik for coordinating this event. We are already looking forward to the next one.

How to stay involved

Follow us on LinkedIn to stay up to date on upcoming events and see what our team is working on.

We're also hiring! If you're interested, or know someone who might be, check out our careers page and submit your resume. Feel free to reach out to Zala with any questions.