TopicPartition podcast

modern data infra is moving beyond the JVM

12/07/2026
0:00
1:17:51
Recuar 15 segundos
Avançar 15 segundos

In this episode, we talk with "Jordan has no life" about a variety of timely data engineering including:- the industry standardizing on columnar processing (eg Arrow) and Rust-based OSS engines- new file format proliferation (Vortex, Lance, Nimble, F3)- object storage as the backbone for modern data systems- the open core business model (e.g the Tabular acquisition and Iceberg's PMC dominance)Jordan is the creator of the popular YouTube channel 'Jordan Has No Life,' ex-Google data-pipelines engineer, now on a data team at a high-frequency-trading firm.We also touched on all the second-order effects from AI coding revolution:- English becoming the working abstraction for code- human reviewers becoming the bottleneck (esp in OSS)- focus on benchmarks, tests and design more than implementation- testing practices for the AI era - triaging flaky tests with AI, fuzz/chaos testing a-la TigerBeetleLast but not least, we touched on Jordan's two latest projects he's been working on: StreamFusion and IceStreamStreamFusion accelerates Apache Flink by rewriting Flink SQL operators in Rust and offloading columnar SQL-plan nodes to embedded DataFusion instances running on each task manager- StreamFusion: https://github.com/datafusion-contrib/StreamFusionIceStream addresses the Iceberg upsert problem by doing change-data-capture on an Iceberg table, maintaining a secondary index in Apache Paimon and asynchronously converting write-optimized equality deletes into read-optimized positional deletes- IceStream: https://github.com/jordepic/icestream--------------------------------------------------------------------*TIMELINE*0:00 Opening0:17 the future of the Data Ecosystem: including new file formats, columnar batching & object storage8:00 Jordan's Background (HFT trading firm, Google)13:21 OSS vs proprietary systems20:09 Jordan's Personal Projects22:50 the future of engineering \w AI29:03 the asymmetry of code review & testing (\w AI again)33:49 jordan's StreamFusion project - accelerating Flink in Rust with DataFusion and Arrow50:16 jordan's IceStream project and the Iceberg Upsert Problem57:52 table formats (Apache Paimon & more)1:04:27 Iceberg examples in prod and the need for real-time streaming to it-----------------------------------------*JORDAN*Jordan is the creator of the YouTube channel 'Jordan Has No Life' (@jordanhasnolife5163), which has several hundred videos on distributed systems and system-design interviews. He has recently pivoted toward more hands-on data-infrastructure projects.He previously worked as a software engineer at Google on marketing-data pipelines. For the last ~2 years he has been on a data team at a high-frequency-trading firm doing offline analysis on self-hosted/on-prem infrastructure (Kafka, Spark, Flink, Iceberg)You can find Jordan on:- Github: https://github.com/jordepic- Linkedin: https://linkedin.com/in/jordan-epstein-69b017177- Youtube: https://www.youtube.com/@jordanhasnolife5163-----------------------------------------*TRANSCRIPT*Feed this into your favorite AI for summarization, or to prompt it specific questions:https://gist.github.com/stanislavkozlovski/5f786853b0a0f248d08d5edb42117635(or just send Gemini this video link and ask it)-----------------------------------------*OTHER PLATFORMS*YouTube:https://youtu.be/5ybQIFb9hE8Apple:https://podcasts.apple.com/us/podcast/topicpartition/id1814926834General RSS:https://anchor.fm/s/104fd76e0/podcast/rss-----------------------------------------If you found anything useful from this episode, please consider supporting our growth (so we can continue delivering valuable content).You can do this by simply liking the video. It takes 2 seconds to do, and recording/producing this takes us 8hrs+

Mais episódios de "TopicPartition"