Databricks User Group Belgium

Join us for the 11th edition of our meetup, filled with insightful sessions on Databricks, great company, and drinks & bites. As always, expect the same great format: a food & drinks reception, two sessions, and a laid-back closing drink.

📍 Location:
Datasense
Corda Campus, Corda 2, Floor 1
Kempische Steenweg 293/16,
3500 Hasselt

🚅How to get there?
The event venue is easily accessible both by car and train. If you’re traveling by train, please get off at Kiewit station. The Datasense offices are located on the first floor of the ‘Corda 2’ building.

🕡 Schedule:

• 6:30 pm: Welcome

Doors open, grab a drink and a bite as we get started!

• 7:15 pm - 7:45 pm: Beyond Dashboards: Deploying a 'Lakebase' App for rule-based Fraud Detection

Starting from the original business case, we’ll explore how we built a scalable fraud detection platform using Databricks Lakehouse, aligning it with our data product strategy. We’ll cover key aspects like data ingestion with Lakeflow Connect, deploying production-ready logic with Databricks Asset Bundles (DABs), and building an interactive investigative tool with a custom "Lakebase" app.

By Ferre Janssen, Data scientist & Engineer @ Datasense
Ferre is a Data Engineer & Scientist with a background in Applied Computer Science. Since 2021, he has specialized in building modern data platforms using Python, Spark, and SQL. His current focus is on rolling out the Databricks platform to build and deploy scalable, production-grade data applications at Eurostar International.

• 8:00 pm - 8:30 pm: Beyond the Limits: Scaling Data Ingestion with Databricks Auto Loader and File Events

In this session, I’ll share my first-hand experience of pushing Databricks Auto Loader to its limits, and what happened when I broke them. We’ll look at how my team built a custom ingestion solution to overcome cloud provider file event quota limits, the architecture behind it, and the challenges we faced. The story then takes a turn with the release of Auto Loader’s File Events mode, a Databricks-managed solution introduced after our work that solved the same problem. I’ll compare our custom approach with this new mode and explain why it should be your default choice for building scalable ingestion pipelines.

By Rob Janssens, Data Engineer @ Cegeka
Rob is a Data Engineer who transitioned from the world of academic research, holding a PhD in Accounting, to pursue his passion for building with data. For the past two years, he has been creating data solutions using Python on the Databricks platform. He is deeply interested in the intersection of data engineering, data architecture, software engineering, and DevOps, always seeking to build not just functional, but also valuable, data products.

• 8:30 pm - 10:30 pm: Networking & Drinks
  End the night with casual drinks and great conversations!