Data Lakehouse
4 blogs ยท 2 talks
The lakehouse architecture: open table formats on cloud storage, and the designs that make them serve both analytics and AI.

How to Save Apache Iceberg from Chilly Meltdown When Dimensions Too Hot to Handle (opens in a new tab)
Fast-changing dimension tables expose bottlenecks in Iceberg's design. How Apache Hudi, built for mutable writes, handles them instead.
Apache HudiApache IcebergData LakehouseSeptember 26, 2025onehouse.ai
Building a RAG-based AI Recommender (1/2)
What retrieval-augmented generation is, and why fresh, well-structured lakehouse data with incremental processing is the foundation of a good RAG app.
Apache HudiAIData LakehouseIncremental ProcessingJuly 11, 2025
Apache Hudi: From Zero To One (10/10)
Highlights of the Hudi 1.0 release that takes the project from zero to one: the LSM timeline, non-blocking concurrency control, and new indexes.
Apache HudiData LakehouseConcurrency ControlIndexingApril 13, 2024
Apache Hudi: From Zero To One (1/10)
A first look at how Apache Hudi lays out a table on storage: the timeline, metadata files, base and log files, and how they fit together.
Apache HudiData LakehouseAugust 28, 2023
Talks

Apache Hudi: The Path Forward (opens in a new tab)
Alluxio Day VI 2021

