Right-time data for enterprises. Estuary replaces fragmented data stacks with one platform for CDC, streaming, batch & pipelines.
New York
- Tweets 513
- Following 42
- Followers 308
- Likes 201
A thousand typical queries at 100 GB cost $0.31 on one warehouse and $7.14 on another. About 23x.
But if yours sits idle most of the day, what the meter does when nobody is querying matters more.
Idle billing measured for all 7 engines in a free report: go.estuary.dev/dwh
1
30
We ran a benchmark of 7 cloud data warehouses and made it open source: 244 queries, TB-scale, up to 100 people querying at once at every possible compute size.
Get the complete guide for free: go.estuary.dev/dwh
3
136
Every new row classified by AI the moment it's written.
@HoytEmerson streams Postgres through Estuary CDC into a live app via webhook, and Jev scores each cart update as it lands. Watch a meal form as items are added.
VideoL piped.video/watch?v=zWY_rXAN…
1
1
104
Run live analytics on your SQL Server data without running them on SQL Server.
New walkthrough on SQL Server CDC into @ClickHouseDB with Estuary, including how ClickHouse handles updates and deletes. Full video in the replies.
Full tutorial here: piped.video/watch?v=qnXXFXve…
2
45
select count(*) from data_planes; → 101 🎉
We just crossed 100 data planes — every place Estuary runs your pipelines in real time, from our own cloud regions to your environment via BYOC.
Your data never leaves your cloud. Here's to the next 100.
1
193
Estuary retweeted
Change Data Capture is simple at its core, but hard to implement properly. CDC tools make this process much easier.
I put a video together showing a simple pipeline of local Postgres -> Estuary CDC -> MotherDuck.
Go deeper here: piped.video/_tgiCopf8aY?si=QFki…
1
4
22
1,545
Estuary retweeted
Change Data Capture is a hard problem to solve correctly. The edge cases are limitless. I've seen some terrible in house attempts to build a solution. I've used Protocol Buffer -> PubSub pipelines before and it's still difficult.
I'm partnering with @EstuaryDev because it challenges me to understand this topic more and their solution is also simple + interesting. I'm a sucker for easy solutions to hard problems.
piped.video/_tgiCopf8aY?si=Yvld…
1
2
456
Meet the new Estuary.
We rebuilt the engine underneath the platform. New transaction loop, written in Rust. Data staged on disk instead of in memory. Each task coordinated by a single leader.
8x more throughput on a single worker.
estuary.dev/runtime
1
4
928
Your team wants ClickHouse for real-time analytics, but no one wants to maintain a second pipeline to make it happen. With Estuary's fan-out architecture, you can capture once and serve multiple destinations.
Join us for a live walkthrough on Aug 26: hubs.li/Q04sQ6j20
58
Stop making Postgres moonlight as your fraud analytics engine. Estuary's CDC captures every change once and streams it continuously into ClickHouse, so fraud scoring and compliance queries run in real time without ever touching your production database.
hubs.li/Q04st9Tp0
1
75
Data movement? More like data momentum 😎 This badge recognition from the @g2dotcom Summer 2026 Grid® Report for ETL Tools is a testament to the power of Estuary's growing community of users. Our momentum is your momentum!
1
27
The biggest event involving an Iceberg since the Titanic. Join us on Tuesday, July 28 for a virtual demo of Estuary's data integration platform and our Iceberg materialization connector. Q&A to follow!
Limited capacity, secure your spot today: hubs.li/Q04qKcc00
36
Estuary Agent Skills give your AI coding assistant the Estuary-specific context it needs to build and operate real-time data pipelines.
Works with Claude Code, Cursor, Codex, and 45+ other agents. 37 public skills currently available and more to come!
piped.video/watch?v=lc71ILyd…
117
Agents won't replace data engineers. They'll cut down the context switching so engineers can focus on actual problems. Read how our Agent Skills setup a real-time Postgres replication in under an hour with Claude Code, no prior Estuary experience needed.
bit.ly/4gYtnRi
1
337
Can your AI agent trust the Postgres data it's acting on? In his new article, Dani Pálma (Dev Relations, Estuary) covers the failure modes that hit agent pipelines hardest, and what a real-time data plane between Postgres and your agents should look like.
bit.ly/4vG9djy
30
It's easy to say "use Postgres" for OLTP, but where do you actually host it? In our new guide, we highlight the best platforms by use-case and around real operational decisions like startup velocity, AWS/GCP/Azure alignment, and infrastructure control.
bit.ly/44ncLv8
46
Next Tuesday! Learn how to setup and deploy a ClickHouse data pipeline in Estuary in this fast-paced walkthrough. RSVP today to reserve your spot: bit.ly/4a4sZwx
46
Two great events in one evening! Last night we hosted a private dinner at The Cavalier to celebrate the launch of our Agent Skills with new and familiar faces.
We also turned up the heat w/ our pals at @orchestra_data for a spicy data happy hour.
Thanks to all who attended!
ALT A welcome sign on a table introduces a celebration for right-time data infrastructure in AI production, mentioning Estuary's new Agent Skills and unlimited throughput updates. The sign includes an illustration of clinking champagne glasses.
ALT A dining table at a restaurant is set with a white tablecloth, black gift bags, and a sign holder displaying "Table Agent Ready." A name card on the table reads "ESTUARY." A fork is placed to the left of a white plate, and a blue bag is visible in the background.
ALT A social gathering in a large, elegant room with chandeliers, tall windows, and curtains. People are mingling and chatting near a white bar with various beverages. The atmosphere appears lively and relaxed.
ALT A group of five people standing together in an elegant room with large windows and dark curtains. There are chandeliers, ornate wall decorations, and a mirror. Some people are holding drinks.
26
Your data is your biggest advantage when building agentic AI infrastructure for your company.
In this new guest post from Mark Van de Wiel (former CTO, Fivetran), he outlines why real-time data is an imperative for effective Snowflake AI architecture.
bit.ly/4vubT3v
30