Date: 06.10.2026, Bucharest, Romania
Databricks Bucharest User Group #3: A Full House at The Marmorosch
On September 30, more than 110 people filled The Marmorosch Bucharest, Autograph Collection, for the third edition of Databricks Bucharest User Group. As proud co-organizers of the largest Databricks meetup in Bucharest, we wanted to say thank you first: to everyone who came, to the speakers who shared real work, and to the Databricks community that keeps this event growing.
The format didn’t change. Practitioners from Romanian organizations talked about what they actually built on Databricks, and the audience asked questions. Three sessions, three industries, one room.
AI-powered energy trading at OMV Petrom
Michael Reisinger, Digital Trading Manager, and Andrei-Cosmin Ciocan, Data Engineering & AI Technical Lead, both at OMV Petrom, opened the evening with a session on how the company uses Databricks for AI-powered energy trading. It was a look at how data and AI feed into trading decisions, from both the business side and the data engineering side, presented together by the two people closest to it.
Migrating the biggest EDW in CEE to Databricks
Mircea Idriceanu, Head of Data for Group; Data Analytics & AI Division at BCR, and our CTO Alexandru Puiu walked the audience through the migration of the biggest enterprise data warehouse in Central and Eastern Europe onto Databricks. Moving a legacy EDW of that size while a bank keeps running is a different kind of problem from a greenfield build, and having the client and the delivery partner on the same stage made it easier to talk about the decisions and tradeoffs from both sides.
Governance: from cost optimization to AI transparency
Vlad Mihalcea, Lead Data Engineer at mindit.io, and Burak Seyid, Solutions Architect at Databricks, closed the program with a session on governance. Their starting point is that keeping compute costs under control and keeping AI workloads transparent and auditable show up together as a Databricks deployment grows, so they are better handled as one conversation than as two. The session ended with the selfie that closed the night.
What a room like this is for
Slides and recordings can only do so much. The value of a user group is the part that happens around the talks: an engineer from one company asking an architect from another how they handled the same problem, a team that is only evaluating Databricks meeting a team that already runs it in production. With a full house, there were a lot of those conversations.
For us, this is also why we co-organize the meetup instead of just showing up to it. We work with Databricks every day, and we want a place in Bucharest where people who work with it can meet, compare notes, and hear honest accounts from the teams who did the work.
Thank you
To Mircea Idriceanu, Michael Reisinger, Andrei-Cosmin Ciocan, and Burak Seyid for sharing your work with the community. To Alexandru Puiu and Vlad Mihalcea for representing mindit.io on stage. To The Marmorosch Bucharest for hosting us. And to the 110+ people who came, asked questions, and stayed to talk afterwards.
We’re already looking forward to the next edition. Follow mindit.io on LinkedIn to hear about it first.
Read more about Databricks + mindit.io:
Case Study: Building customer matching on Databricks for a luxury brand selling worldwide
Databricks partner page: Stop experimenting with AI. Go from legacy estate to AI-native, on Databricks.
Webinar: Databricks Just Moved the Goalposts. Here’s What That Means for Your Team.
