Database Seminar - Vinoth Chandar

— 5:30pm

Location:
Virtual Presentation - ET - Remote Access - Zoom

Speaker:
VINOTH CHANDAR , Founder and Chief Executive Officer
Onehouse
https://www.linkedin.com/in/vinothchandar/

Apache Hudi: A Database Layer over Cloud Storage for Fast Mutations and Efficient Queries

Data lakes emerged as a way to store vast amounts of data as files and objects on infinitely scalable cloud storage, with processing done on scalable distributed compute engines. However, this architecture lacks many of the capabilities of traditional databases, such as efficient mutations, indexing, and transaction management. Apache Hudi was created as the first "lakehouse" project, to bridge this gap by introducing a database-like abstraction on top of file-based data lakes.

This talk will explore Hudi’s design choices and tradeoffs across metadata management, indexing, storage layout, and concurrency control—decisions that enable fast incremental reads and writes while significantly reducing processing costs and query latency. We will also share practical guidance for using Hudi effectively in modern data platforms and highlight open challenges the community is actively tackling, from scaling metadata to supporting emerging AI and unstructured data workloads.

—

Apache  is the original creator of Apache Hudi, a system that brings database-like primitives on top of data lakes. He is the founder and CEO of Onehouse, where he focuses on making lakehouse infrastructure open and cost-effective. Previously, he was principal engineer at Confluent working on Kafka/ksqlDB. He led the data architecture during growth years at Uber and also lead engineer on the Voldemort key-value store at LinkedIn. His work spans distributed storage, stream processing, and real-time data infrastructure. 

This talk is part of the Future Data Systems Seminar Series.

Zoom Participation.  See announcement. 

For More Information:
db-www@cs.cmu.edu

Add event to Google
Add event to iCal