Budapest Databricks Meetup #4: AI Agents & AI Coworkers in Practice
The Budapest Databricks Meetup returns on October 2 for its fourth edition. Organized by Abylon Consulting, the event will once again bring together Budapest’s Databricks, data, and AI community for an evening of practical insights, knowledge sharing, and professional networking.
This time, the focus is on AI agents in practice. We’ll explore two different sides of how AI is changing everyday technical and data work: how multiple coding agents can work together as part of a single workflow, and how Databricks Genie One can help turn enterprise data into an AI coworker that business users can actually rely on.
Both sessions will take a practical look beyond the technology’s promises, with real use cases, live demonstrations, lessons learned, and the challenges teams need to consider when putting these capabilities into practice.
Omnigent: Stop Babysitting Your Coding Agents
Presentation by Oleksandra Bovkun, Senior Developer Advocate at Databricks
“Human-in-the-loop” is supposed to mean that a person makes the judgment calls a model can’t. In practice, when Claude Code, Codex, and Cursor are all open at once, it can mean something much simpler: copy-pasting the same error into different chat windows, re-explaining the same context each time, and losing track of which agent has already tried which fix. Whether you’re debugging an app or a data pipeline, the challenge is the same: the human in the loop is you, and it can quickly become exhausting.
Omnigent addresses a simple question: what if that were one loop instead of three? It is an open source meta-harness that sits above Claude Code, Codex, Cursor, and other coding agents, allowing you to run them as one workflow, swap agents without rewriting everything, set spend caps and approval policies, and share a live session that a teammate can review or co-drive.
In her session, Oleksandra will explore what this looks like in practice, including how agents can communicate and debate with each other, how usage policies work, and how to set up the workflow effectively. Expect a live demo, lessons learned, and practical tips for getting the most out of Omnigent.
Turning Our Lakehouse into an AI Coworker: How to Use Genie Effectively in Everyday Work
Presentation by Tamás Kakuk, Solution Architect & András Kovács, Senior Solutions Architect at Abylon Consulting
Genie One is where enterprise data learns the language of the business.
Powered by Genie Ontology, it transforms disconnected tables, documents, metrics, and applications into a connected understanding of business concepts, entities, and relationships. Genie One uses this context to provide governed, relevant answers, automate workflows, and help users move from questions to insights, and from insights to action.
In this session, Tamás and András will explore how Genie One can support everyday work and examine what lies behind its promises. They will look at practical use cases, the data and governance foundations it requires, and the challenges that can prevent it from becoming a reliable AI coworker.
Want to learn more about Databricks Genie? Read our article on Databricks AI/BI and Genie.
Who is this Databricks Meetup for?
The event is particularly relevant for data engineers, developers, data architects, BI professionals, and AI practitioners, but it is open to anyone interested in Databricks, modern data platforms, and practical enterprise AI.
If you work with Databricks, experiment with AI coding agents, or want to understand how Genie can make enterprise data more accessible, join us. After the presentations, there will also be plenty of time for networking and professional discussions with the local data community.
If data, innovation, and community matter to you, don’t miss out! Join the Budapest Databricks Meetup Group and become part of our professional community.
