Matt Topol

Time for @oredev !! If you're here don't miss my talk today at 10:10 on using @ApacheArrow with #ml workflows! Looking forward to a day of interesting talks and discussions

@andrewlamb1111 @ApacheDataFusio @ApacheArrow Well, most of the systems...

@andrewlamb1111 @ApacheDataFusio You know what's missing from that diagram? The fact that @ApacheArrow is the data format used to communicate between all those systems given that cuDF and Rapids are built on top of the Arrow format...

Since everyone else is feeling the love from Nvidia today, I would also like to point out that @ApacheDataFusio is up there as well ❤️

@_Felipe The 9th is a Thursday lol I'll be in SF all that week though, from Monday 6th till Friday the 10th. We should definitely meet up though!!

@zeroshade I arrive early Monday. The 9th! Let’s meet!

I will be in San Francisco in April and I’m prepared. I can “well akshually” these translations. That’s how prepared I’m.

The fastest operation is the one you don’t have to do. When a database natively supports @ApacheArrow, ADBC can speed up fetching and ingestion by eliminating costly row/column conversions. How much faster is it in practice? We ran some benchmarks to find out. Link below 👇

If you aren't paying attention to some of the Apache Spark Acceleration projects like Gluten, you should! Gluten just graduated as a Top-Level Project @TheASF

Fetch query results without ODBC / JDBC bottlenecks. The new ADBC driver for @databricks is now in early release. Install it with dbc. Details in comments.

私の稼働が空いちゃったので、私と一緒になにかお仕事したい人からの連絡を待っています！ 須藤、空いています - 2025-11-27 - ククログ clear-code.com/blog/2025/11/2…

🚀 Launching Supermetal — data replication that just works. Sync databases to warehouses in real-time or batch — no Kafka, no JVM, no Debezium. Built in Rust & Apache Arrow. Try it → trial.supermetal.io Launch post → supermetal.io/blog/launch #dataengineering #rustlang

Data and AI are evolving fast, but much of today’s infrastructure still runs on standards from the 90s. @columnar_tech, from the team behind Apache Arrow, is bringing an Arrow-native protocol (ADBC) that moves data 10–100× faster across systems like Snowflake and DuckDB. We're excited to lead Columnar's $4M seed round. Read the full Q&A to learn more: bessemervp.team/47d8gWk

A new startup, @columnar_tech, looks to streamline the copying of tabular data across systems, using @ApacheArrow and the ADBC API. By @Joab_Jackson thenewstack.io/apache-arrows-…

Come join the community with our launch! Learn more and come talk to us about data connectivity with @ApacheArrow !!

ODBC is getting tired. It can't keep up with the fast new kids in the data world these days. The next generation is ready to take the torch. Meet ADBC, a fast, modern data connectivity standard built on @ApacheArrow. Watch my talk from the @CMUDB seminar: youtu.be/TjlmNGNx77E

We're building the data infrastructure that AI actually needs. Current systems were built for humans reading dashboards. But an H100 can consume 4 million images per second. The future isn't human-scale. It's machine-scale. Introducing Spiral: Data 3.0 🌀 1/8

Its happening -- DataFusion will (finally) get spilling hash joins. The march to completeness begins

In September the @columnar_tech crew are headed to @PyDataParis 2025 and the first ever @ApacheArrow Summit. The organizer @QuantStack is a dedicated supporter of Apache Arrow. We’re delighted to be sponsoring the event.

Excited to see continued improvements in embedded query processing in the @ApacheArrow C++ project: arrow.apache.org/blog/2025/07/1…

1/5. @paradedb has raised a $12M Series A to bring Elasticsearch workloads to Postgres. 🧵 techcrunch.com/2025/07/15/par…

@miguelperedo @mim_djo @haro_ca_ True, though I'm one of the people working on getting that support working all the way into iceberg :)

@zeroshade @mim_djo @haro_ca_ Yes, but for example not yet in Iceberg github.com/apache/iceberg… . So I just need to try if it works in duck lake …

#DuckLake first impression Good: "First Python Engine(not Spark), where you can create tables and it writes Parquet files, and it has proper SQL support." Bad: "Parquet positional deletes instead of Data Versioning (DV)." Ugly: "All files in the same folder, not per table."