

DS Talks: Building AI Agents That Actually Work, with Larsen Cundrič
Join us for our next DS Talks event with Larsen Cundrič, a Slovenian engineer who went from studying computer science in Ljubljana to Silicon Valley.
AI agents are everywhere right now, but a capable language model is only the start. Making agents reliable, safe and scalable is where the real engineering happens. Larsen will share what he has learned from building AI agents and AI products in a fast-moving startup environment.
What we’ll cover:
How AI agents work: the fundamentals of agents that interact with browsers, tools and external systems
The engineering challenges: reliability, sandboxing, infrastructure and scaling, and why a useful agent needs much more than a good model
Building AI products: lessons on engineering and development practices when the field changes every few months
From Ljubljana to Silicon Valley: Larsen’s career path, the opportunities out there, and advice for everyone who wants to get into the field
The talk focuses on general engineering principles, publicly available technologies and personal experience rather than company-specific implementations.
We’ll finish with an open Q&A and of course, refreshments after that.
Whether you’re a student, an engineer curious about AI agents, or thinking about a career in the US tech scene, this one is for you. See you there!