The traditional approach of “AI is just a tool” limits AI to a supporting role, is outdated, and seems outright amusing in the context of the emerging boom of AI agents. Fully unlocking the potential of modern AI systems requires interactive, two-way collaboration and partnership. Let’s explore various models of human-AI interaction that are transforming work paradigms: the “cyborg” and “centaur” models, the “ensemble,” the “author-driver” and “resonator,” the “cybernetic colleague,” the “thinking partner,” and others. How can we avoid the “falling asleep at the wheel” effect or disengaged AI interaction? Every employee becomes the CEO of a startup composed of a mini-team of AI agents — but what is the optimal number of AI agents per employee?

I’ll talk about how the approach to writing changes when a generative AI becomes your co-author. I’ll share my personal experience of writing a book about generative AI — with its help — including ethical dilemmas, cognitive challenges, and unexpected insights. I’ll explain why the process took ten times longer but brought a hundred times more experience. I’ll also show how I worked with terminology and how, in parallel, I developed practical recommendations for using generative AI in programming, education, and other fields.

I will discuss modern video generation capabilities from text prompts and AI-assisted editing, leveraging my 17 years of video production experience. I will showcase key platforms and tools, share case studies of integrating AI into creators’ and product managers’ workflows, outline product opportunities and workflows for developers and video makers, and highlight priority directions for the development of AI-based video services.

Can you trust the cloud? Confidential Computing enables data protection even in fully controlled environments.

A practical case of using a computer vision system for automated visual inspection in a pharmaceutical production line. The system ensures the quality and consistency of vial contents, reducing human error and improving efficiency in a highly regulated environment.

As an infrastructure provider, we often see that potential clients are somewhat disoriented when it comes to choosing the right accelerators. The technology evolves rapidly, and questions like “What’s out there?”, “How do these cards differ?”, and “Which ones are better?” are completely valid. But the selection criteria are far from simple. In this talk, we’ll explain key concepts such as compute-bound vs memory-bound models, arithmetic intensity, and model quantization—factors that are crucial when choosing the right hardware for your workload.

Kernel is currently the leading producer of sunflower oil and one of the largest agroholdings in Ukraine. What business challenges are they addressing, and why is ML a must-have? This talk explores the development of the data science team at Kernel—from early experiments in Google Colab to building minimal in-house infrastructure and eventually scaling up through an infrastructure partnership with De Novo. The session will highlight their work on crop yield forecasting, the positive results from testing on H100, and how the speed gains enabled the team to solve more business tasks.

A talk by an AI solutions integrator on how they structure their internal business processes and tasks using AI. Learn how a model can process business requirements in just 15 minutes—something that would typically take a data science team two weeks to complete.

Let’s explore Computer/Browser/Mobile Use agents. We’ll start with the APIs provided by OpenAI and Claude that support such use cases. Then, we’ll recall how LLMs and VLMs are trained, what Reinforcement Learning (RL) is, and how it can be applied in this context. We'll also look into some recent open-source agent models, and discuss how to evaluate these agents effectively.