Thrilled to share that I'll be talking about using #ApacheFlink for data enrichment at #Current24 in Austin, Sept 17-18: confluent.io/events/current/ Hope to see you there!

Looking for practical #ApacheFlink use case examples? Our experts, David Anderson, @MartijnVisser82, and Chesnay Schepler, have gathered a collection of recipes 🧑‍🍳 to help you tackle a variety of common on-premise challenges. Explore the resources: cnfl.io/github-flink-c…

❌ Delete the duplicates! With exactly-once semantics, each message is delivered precisely once. But how does it work? Learn about the concept and how #ApacheFlink enables exactly-once processing for your real-time streaming data with our new video: brnw.ch/21wJy4U

👨‍🍳 Looking for practical #ApacheFlink use case examples? We prepared a cookbook with a collection of best practices, with real-life use cases including: 🌰 Joining and deduplicating data 🌰 Deserializing JSON from #Kafka Plus many more! Start cooking: cnfl.io/github-flink-c…

It’s no secret that LLM training data is running out. How close are we to the limit? To answer that, here's an estimate of the total amount of text in the world from every major source:

On July 25th, we're inviting #apacheFlink to our #apacheKafka party. Hear from David Anderson and Sharon Xie. meetup.com/kafkabayarea/e…

I've spent the past ~3 weeks going through the entire history of deep learning and reimplementing all the core breakthroughs. It has completely changed my beliefs about deep learning progress and where we're headed. Progress tracker in thread (all resources at the end) 👇

1/ There is a mini open source drama, this time because RedPanda, a company with a source-available Apache Kafka clone, bought an open source connector framework and made licensing and trademark changes to thwart other startup competition. (a thread) news.ycombinator.com/item?id=405379…

Adding queue semantics as an API over logs provides the cooperative consumption advantages of traditional queues while fixing some of the queue structure's disadvantages: lack of replay and the need for write amplification. Read more in Jack Vanlightly's blog: cnfl.io/449EFd0

Our pre-#current24 summer meetup series continues on July 25th with some #apacheFlink! 🐿️ Sign up now to catch Sharon Xie and David Anderson as they dive into timing in #flinkSQL and enrichment patterns. ⤵️ meetup.com/kafkabayarea/e…

Blogged: Predicting the Future of Distributed Systems blog.colinbreck.com/predicting-the…

Meet the new emerging role: The Data Streaming Engineer. During our day 2 #Current24 Keynote, learn how this role is bringing together the work of app developers, data engineers, data scientists and more—potentially redefining the work you already do.

He’s right! David Anderson and I can’t stop, won’t stop practicing for #current24.

You haven't seen complexity in art until you've seen this 600-year-old painting. It's so detailed that modern medicine has been able to diagnose this man's exact type of blindness. And that's where things get mind-bendingly strange — this whole thing is an illusion... 🧵

Watermarks are at the heart of what makes stream processing with #FlinkSQL possible. Join David Anderson as he explains the ins and outs of watermarking in in the latest video from his new course on #ApacheFlink SQL. Check it out on Confluent Developer ➡️ cnfl.io/3XSiR2Z

It's not too hard to build a stream processor, right? Operators, network, add checkpoints, done. Making state and snapshots work really well and efficiently has been a continuous effort since we introduced this in Apache Flink in 2015. Here is a nice article revisiting all that

Prequal is a must-read paper that makes YouTube +5-10% faster. They show that load balancing w/ requests-in-flight and latency outperforms weighted round robin on CPU load (and why). A rare paper showing distributed systems research working in production. Here’s how: 1/4

I have a new video up on YouTube — youtu.be/5TK8UrmU83I — about doing complex event processing with #apacheflink. This might sound like an obscure topic, but I have found this technique to be super-powerful, and broadly applicable.

Yesterday we hosted SF's most exclusive demo day. Only 10 teams, all of them with moonshot visions. Here's an inside look at the demos: (🧵)

We’ve had a blast building this. Hope you like it!