AI systems in practice

Doors open, registration, drinks & snacks

The rise of multimodal, agentic AI

Shai Alon is back in Warsaw, and he's bringing the future of AI with him! This session explores a paradigm shift in artificial intelligence as we move beyond traditional, single-input models to proactive agents that can process and respond to various inputs – text, images, audio, and real-time interactions – while anticipating needs and taking initiative. Featuring demos showcasing Gemini 2.5's multimodal capabilities, you'll leave with a clear understanding of the concepts, practical tips for getting started, and inspiration to build your own multimodal, agentic AI systems.

Coffee, beer, networking

‍

Cloud? No thanks! I'm gonna run GenAI on my AI PC

This presentation introduces the AI PC, a single machine capable of hosting diverse AI applications that run GenAI in seconds, not hours. Learn how local AI overcomes common issues with Cloud AI, including data privacy risks, high latency, and cloud dependency. The session showcases the OpenVINO Toolkit, an open-source solution for optimizing and deploying deep learning models to maximize AI performance on your PC. Live demonstrations include people detection and image generation, with all presented demos available on GitHub.

‍

Surviving the Wild West of AI in 2025

A brutally honest overview of the AI landscape in 2025. Where are we really — and why did we end up here? This session goes beyond glossy, optimistic visions to explore what's happening beneath the polished headlines and marketing promises. Through raw examples, the presentation examines the risks, failures, unintended consequences, and uncomfortable truths most prefer to ignore. No dark corner left unexplored, this is your essential survival guide for navigating the chaotic, unregulated Wild West of AI in 2025.

Coffee, beer, pizza, networking

‍

The state of AI security: threats, mitigations, and trade-offs

Design clusters for AI workloads — the hard way

Daniel shares practical insights and hard-earned lessons on architecting and building infrastructure for AI workloads—from scrappy startups to large-scale enterprise systems. This session walks through best practices, common pitfalls, and what he learned the hard way so you don't have to. Learn practical strategies for designing resilient, scalable clusters that handle the unique demands of modern AI applications while avoiding costly mistakes.