Kraków AI | 2nd Meetup

Join us on 2nd Kraków AI Meetup @ Beekeeper!

The community-driven event dedicated to exploring the vast potential of Generative AI and general ML. Whether you are an Engineer, Researcher, Product Manager, or simply an AI enthusiast, this Meetup is designed to inspire and give you an opportunity to share back.

📅 Date: 4th March 2025 (Tuesday)
🕕 Time: 6:00 PM
📍 Location: Beekeeper office - V.Offices Al. 29 Listopada 20 Kraków
🍕🍻 Pizza and drinks provided
🔗 Do not hesitate to RSVP!

This time, we are trying something new! We are hosting a live-streamed meetup from our Zurich office, and we would love for you to join us in person at our Krakow office to be part of the experience.
You will have the chance to ask live questions, engage with the speakers, and, as always, enjoy some great networking opportunities.

Agenda:
Valery Lezhebokov (Coresystems AG) – Building an AI-powered knowledge base for service (field, customer, self-service, etc.)
Luca Perrozzi (AWS) – From Bedrock Agents to Model Context Protocol: What? Why? How?
Michał Koźmiński (Beekeeper) – ClickHouse as the Foundation for Scalable and Drift-Resistant AI Predictions
Abstract:
Valery Lezhebokov will discuss building an AI-powered knowledge base to enhance service efficiency. The talk will cover key architectural and technological decisions made for scalability and performance, Valery will explore deployment strategies, comparing Large Language Model solutions available as PaaS versus self-hosted models, providing attendees with practical insights for their AI projects.
Luca Perrozzi explores Bedrock Agents and Model Context Protocol architectures, functionality, and implementation patterns for building AI applications with structured reasoning capabilities.
Michał Koźmiński explores ClickHouse as a robust foundation for scalable and drift-resistant AI predictions. The presentation emphasizes how materialized views streamline feature preparation, ensuring accurate and interpretable predictive outcomes within data pipelines. Attendees will gain valuable insights into effective strategies for monitoring and mitigating drift, maintaining model stability, and consistently delivering high-quality predictions.