DevBP #17: Rebuilt & Refactored

In the upcoming DevBP meetup, we will venture into the world of core architecture, exploring the craft of outgrowing legacy abstractions - from replacing fragile 3D algorithms with clean spatial pipelines, to evolving early Big Data setups into modern, Iceberg-powered lakehouses.

=== === === ===

A Story of a New Support Algorithm - Dmitri Megretski [Formlabs]

When 3D printing complex models, temporary supports are required to hold up overhanging geometry. Formlabs software relied on four conflicting algorithms to place these supports, causing slow performance and unpredictable results. This talk covers how Formlabs rebuilt the system into a predictable two-stage pipeline: first mapping out where a model needs structural help, then running an optimized search algorithm to place supports cleanly. Learn how the team eliminated geometric precision errors by evolving their internal data representation from basic pixel grids to sparse 3D voxels.

=== === === ===

Modern Big Data with Apache Iceberg - Attila Turoczy [Cloudera]

Early Big Data architectures struggle with modern engineering demands like schema evolution, row-level updates, and transactional consistency. This talk traces the shift from traditional Hadoop setups to modern lakehouses, exploring how open table formats like Apache Iceberg bring standard database reliability to cloud-scale data analytics.

=== === === ===

This English-language event will be held at the Formlabs Budapest office, where we'll provide snacks and drinks to fuel conversations.

Arrival is from 17:30, the presentations start at 18:00.

You can also find the recordings of previous events on the @DevBPMeetup YouTube channel.