​🚀 Connect@Lingaro - Databricks meetup (Warsaw)

​This time in #collaboration with the Databricks Warsaw community with special guest from Databricks - Kiryl Halozhyn.

​📅 November 18 (Wednesday)
⏰ 17:30–21:00
📍 Lingaro Office, Puławska 180 Warsaw - 4th floor.

​It will be an evening of practical insights, experience sharing, #TechExcellence conversations and Polish pierogi 🥟.

🎤 Speakers & topics:

​Kiryl Halozhyn, Field Engineering Leader @ Databricks

​He leads a team of Solution Architects helping enterprise customers take Data & AI from platform to production. Previously a Senior Solutions Architect at Databricks and a Solutions Engineer at Cloudera, he has worked across Hadoop and lakehouse platforms in the last 10 years.

​"What's New in Genie One - From Chat to a Multi-Agent Coworker"

​Genie started as conversational analytics. In 2026 it became Genie One: a data-smart AI coworker with a living context layer, reusable skills, governed tools, and domain agents you can create from a conversation.

​This session is for engineers and architects. We will skip the “chat over tables” recap and look at the current product surface: how Genie Ontology infers and ranks business context from Metric Views, dashboards, SQL (including dbt/query history), and Agents; how skills published through Unity Gateway become shareable capabilities; and how MCP, a code-execution sandbox, scheduled tasks, and “agent from chat” change the architecture of internal automation. I will show how we actually use this at Databricks and what is upcoming soon in the product (quite a few cool things!). You will leave with a mental model that sits next to context engineering for coding assistants: how to give production AI a governed runtime over enterprise data.

​Łukasz Baczewski, Data Architect @ BEC Financial Technologies

​As a Data Architect with a strong interest in cloud data architecture, Databricks, data products, and data governance. He is particularly interested in how architecture, security, lineage, and data segregation come together to create scalable and sustainable data solutions.

​"When Agents Touch Sensitive Data"

​Data access controls answer “who can read this table”, but AI agents raise a different question: what happens to the data after it’s read? Once sensitive information enters an agent’s context, it can end up in a summary, a message or a report sent to the wrong audience. This talk explores how data classification, agent-level policies and platform-level guardrails can each help reduce that risk, and where each of them falls short.

​Marcin Kostkiewicz, Karol Mierzejewski & Bartłomiej Henzel - Data Engineering team from Lingaro

​"Give AI a Brain: Building a Shared Semantic Foundation for AI Agents and Databricks Genie"

​AI can write code, generate SQL, and answer business questions. But does it truly understand your data platform - its architecture, data flows, business concepts and metrics?

​As AI becomes an integral part of Data Engineering and Analytics, the real challenge is no longer just choosing the right model. It's building the right context and semantic foundation behind it.

​In this session, we'll explore two sides of the same challenge.

​Part 1 - AI for Data Engineers

​Discover how architecture documentation, C4 models, code-derived data lineage, ontology, coding standards, and agent-driven workflows such as BMAD can transform an existing Databricks project into an AI-ready development environment.

​We'll show how structured knowledge helps AI agents navigate complex data pipelines, understand dependencies, and implement features consistently with existing engineering standards.

​Part 2 - AI for Data Analysts

​We'll take the same knowledge beyond the code and into Databricks' evolving semantic ecosystem: Unity Catalog, Metric Views, Domains, Pages, Genie Ontology, Genie One, and Genie Code.

​Explore how business definitions, trusted metrics, and data relationships can provide AI with the context needed to generate more accurate, consistent, and explainable analytical answers.

​See you there! 👋🏼