PeerDB (YC S23)

PeerDB (YC S23)

Software Development

Fast, simple and cost-effective Postgres replication. Now part of ClickHouse

About us

PeerDB provides a fast, simple and a cost effective way to replicate data from Postgres to Data Warehouses like Snowflake, BigQuery, ClickHouse and Queues like Kafka, Redpanda, Event Hubs, PubSub etc. PeerDB is 10x faster and 80% cheaper than existing ETL tools for Postgres replication. We are laser focused on Postgres. So if you are heavy Postgres user and are looking for replication solutions, PeerDB can be of value.

Industry
Software Development
Company size
2-10 employees
Headquarters
San Francisco
Type
Privately Held
Founded
2023

Locations

Employees at PeerDB (YC S23)

Updates

  • PeerDB (YC S23) reposted this

    View profile for Benjamin Wootton, graphic

    Independent / Freelance Consultant - Cloud, Data & AI

    This week I spoke at the ClickHouse meetup here in Dubai on how ClickHouse Cloud can strip out the complexity from a modern data stack. One of the reasons I put forward was how PeerDB and ClickPipes (the data integration feature of ClickHouse Cloud) allows you to ingest data directly, potentially avoiding a lot of tech such as Airbyte, Airflow, Fivetran, Kafka Connect and the like. I've recently put out 3 blog posts walking through how this is setup. The first shows how we can use open source PeerDB to replicate data from Postgres to ClickHouse, the second walks through ClickPipes as integrated into ClickHouse Cloud, and the third includes a video demo of the two options for people who prefer that format. The aim was to show how we can easily setup a real time CDC sync with very little effort and ceremony. My belief is at a time when businesses are trying to save money and remove some of the complexity that has built up in their data estate, direct replication processes like this could be a very useful pattern to consider. 3 blog articles linked in the comments!

    • No alternative text description for this image
  • 🎉 We are excited to announce PeerDB v0.24.0, our second release of the year! This release of PeerDB features code changes to improve performance, observability and fine tuning of replication workflows. ⚡ PeerDB now pulls from Postgres and pushes to ClickHouse in parallel (async), by default 🔧 Introduced knobs to tune memory usage of PeerDB inserts to ClickHouse via chunking ✅ Improved validation for read replica Postgres instances below PG16 📊 Added collection of more metrics such as errors emitted and rows synced 🚀 Re-architected parallelized syncs to ClickHouse and single table sync performance Here go the full release notes: https://lnkd.in/gVQFd7VN

    • No alternative text description for this image
  • PeerDB (YC S23) reposted this

    View profile for Michael Driscoll, graphic

    Co-founder at Rill Data, fast dashboards via GenBI. Previously founded Metamarkets (acq'd by Snap) and CustomInk.com. Founding partner at DCVC.

    Whoever said "never meet your heroes" should NOT attend Data Council this April. :) Here are my hot takes for why I'm so excited these legends are gathering together in the Bay Area: - Hadley Wickham created R's tidyverse (ggplot2, dplyr) and is a founding father of data science at Posit PBC - lloyd tabb created Looker, the most successful cloud BI tool ever, sold it to Google for $Bs and is back building a new language for data, Malloy, at Meta - Hannes Mühleisen co-invented DuckDB, the wildly popular open-source database with 10M+ downloads a month - Andy Pavlo, CMU professor and YouTube sensation who makes database theory fun and fascinating - Tanya Bragin, the product genius behind the ClickHouse Cloud rocketship, the best executed cloud database SaaS I've ever used - Pedram Navid, data engineering impresario at Dagster Labs whose writings are often hilarious and always insightful (follow him on X / bsky) - Simon Hørup Eskildsen is leading the data-infra-on-object-storage revolution, building a vector database on S3 at turbopuffer Apologies for the other heroes that are speaking and I didn't give a shout out too. A few of us mere mortals will also be attending / speaking, hope to see many of you there! :)

    • No alternative text description for this image
  • PeerDB (YC S23) reposted this

    🚀 Latest release of PeerDB for Postgres CDC! PeerDB is now fully integrated into ClickHouse Cloud and available in private preview through ClickPipes. Sign up here: https://lnkd.in/gD2z42JV

    View organization page for PeerDB (YC S23), graphic

    2,257 followers

    📣🎉 We are excited to announce the first major release (0.23.0) of PeerDB in 2025! This release focuses on enterprise-grade Postgres to ClickHouse replication, with key features centered on improved performance and enhanced reliability for Postgres CDC. 📦 Granular formats sync of binary and hex data from Postgres to ClickHouse. 🚀 Improved throughput for CDC by avoiding reconnections to support enterprise workloads. 🔁 Improved performance by using multiple replicas for data ingestion in ClickHouse. ⏳ Drastically reduced slot growth on idle databases by frequent acking on PKMs. 📊 Added internal endpoints for additional telemetry, including bytes moved and so on. ✅ Added additional validation to detect connection poolers. 🗑️ Improved reliability for Resyncs by removing capture of soft-deleted rows. 🛠️ Richer datatype support, including UUID arrays and tstzrange. 🔧 Revamped retry logic to better identify actual errors. 🧪 Enhanced end-to-end tests for various cases, such as partitioned tables. Full release notes: https://lnkd.in/geuaXYxZ 🚀

    • No alternative text description for this image
  • 📣🎉 We are excited to announce the first major release (0.23.0) of PeerDB in 2025! This release focuses on enterprise-grade Postgres to ClickHouse replication, with key features centered on improved performance and enhanced reliability for Postgres CDC. 📦 Granular formats sync of binary and hex data from Postgres to ClickHouse. 🚀 Improved throughput for CDC by avoiding reconnections to support enterprise workloads. 🔁 Improved performance by using multiple replicas for data ingestion in ClickHouse. ⏳ Drastically reduced slot growth on idle databases by frequent acking on PKMs. 📊 Added internal endpoints for additional telemetry, including bytes moved and so on. ✅ Added additional validation to detect connection poolers. 🗑️ Improved reliability for Resyncs by removing capture of soft-deleted rows. 🛠️ Richer datatype support, including UUID arrays and tstzrange. 🔧 Revamped retry logic to better identify actual errors. 🧪 Enhanced end-to-end tests for various cases, such as partitioned tables. Full release notes: https://lnkd.in/geuaXYxZ 🚀

    • No alternative text description for this image
  • PeerDB (YC S23) reposted this

    View profile for Sai Krishna Srirampur, graphic

    Building PeerDB - Fast, native data-movement for Postgres

    🚀 We will be actively growing our team in 2025! This role offers a great opportunity to dive deep into databases, data-movement and ETL. Our vision is to make database integrations with ClickHouse magical. You'll also have have an opportunity to work with an exceptional team at ClickHouse / PeerDB (YC S23). Don't miss applying for this role!

    View profile for Kaushik Iska, graphic

    Building the next generation ETL

    We are hiring a Software Engineer to join my Database Integrations team. You’ll have the opportunity to work with a driven, supportive team that has built platforms capable of handling petabytes of data, presented at conferences, and contributed to open-source initiatives. If you think you’d be a good fit—or know someone who is—I’d love to hear from you! 🚀 https://lnkd.in/eryqtQdN

    Senior Software Engineer - Database Integrations

    Senior Software Engineer - Database Integrations

    boards.greenhouse.io

  • PeerDB (YC S23) reposted this

    View profile for Kaushik Iska, graphic

    Building the next generation ETL

    When we were building the parallel initial snapshot, we didn’t have customers with 10+ TB databases yet. But we had this conviction—if we wanted to land enterprises, we’d need something special. We spent w debating if the added complexity was worth it. In the end, there was no question—it absolutely was. It was such an intricate feature—keeping a persistent snapshot connection, partitioning by CTID, and ensuring everything scaled smoothly. Focusing deeply on Postgres and ignoring every tempting distraction to expand early is what got PeerDB to this point. Seeing feedback like this makes all this worth it. 🙌

    View organization page for ClickHouse, graphic

    102,809 followers

    🎉 We acquired PeerDB (YC S23) earlier this year to make it seamless to replicate your Postgres databases to ClickHouse. Below is raw feedback from our customer Unify on their Postgres CDC experience. Thank you, Mitchell Bregman, for the feedback! 🙏 We are excited that you were able to easily integrate your Postgres database with ClickHouse. PeerDB is natively integrated into ClickHouse Cloud and powers the Postgres CDC connector in ClickPipes, which is currently in private preview. You can sign up using this link: https://lnkd.in/eZ4WYpsS

    • No alternative text description for this image
  • PeerDB (YC S23) reposted this

    View profile for Mitchell Bregman, graphic

    Eng @ Unify | Ex-Flock Safety

    Huge fan of the amazing work coming from the PeerDB (YC S23) / ClickHouse team!

    View organization page for ClickHouse, graphic

    102,809 followers

    🎉 We acquired PeerDB (YC S23) earlier this year to make it seamless to replicate your Postgres databases to ClickHouse. Below is raw feedback from our customer Unify on their Postgres CDC experience. Thank you, Mitchell Bregman, for the feedback! 🙏 We are excited that you were able to easily integrate your Postgres database with ClickHouse. PeerDB is natively integrated into ClickHouse Cloud and powers the Postgres CDC connector in ClickPipes, which is currently in private preview. You can sign up using this link: https://lnkd.in/eZ4WYpsS

    • No alternative text description for this image
  • PeerDB (YC S23) reposted this

    View profile for Sai Krishna Srirampur, graphic

    Building PeerDB - Fast, native data-movement for Postgres

    This will be a huge one! A lot of game changing announcements - innovations in the Data Lake space, JOIN improvements, native JSON support and more 🚀🚀. Also, ClickHouse release calls are one my favorite recurring sessions to attend. Very real, filled with live demos and led by ClickHouse creator himself Alexey Milovidov. Highly recommend not miss this one and get to know what we’ve been innovating at ClickHouse!

  • PeerDB (YC S23) reposted this

    View profile for Sai Krishna Srirampur, graphic

    Building PeerDB - Fast, native data-movement for Postgres

    Great discussion on hn on Postgres for everything. https://lnkd.in/gnm7rAUA TLDR; I see a bunch of comments: just don’t do it! I’m glad this is becoming mainstream. Don’t get me wrong—I’m a huge Postgres fan and have spent 10 yrs helping customers implement it. However, I’m a strong believer in using Postgres for what it’s designed for in first-place. Postgres was designed as a OLTP database, with 30+ yrs of effort on making it robust for that use case.I know there are many extensions attempting to make Postgres support other usecases, like analytics,queues etc. Keep in mind that these extensions are relatively recent and aim to retrofit new capabilities to a db primarily designed for OLTP workloads. It’s like adding an F1 car engine to Toyota Camry. Will that work? Extensions also have many issues-they are not fully Postgres-compatible. In Citus for example, we added support for the COPY command 4 yrs into the company, and chasing SQL coverage was a daily challenge for 10 yrs. Being unable to use the full capabilities of Postgres and having to work around many features defeats the purpose of a Postgres extension. On the other hand, you have specialized alternatives like ClickHouse, Snowflake for analytics, Redis for caching and Kafka for queues. These solutions have benefited from decades of development, laser-focused on specific use cases. They are highly efficient for their intended purposes. I often hear that these Postgres extensions are expanding the boundaries of what Postgres. While I partly agree, I question the extent to which these boundaries are truly being expanded. In this era of AI, where data is growing exponentially, handling scale is critical for any tech. These boundaries will likely be broken very quickly. Take queues as an eg.: you have a purpose-built technology like Kafka or a Postgres extension. For an early-stage startup, adopting a less optimized Postgres solution may (not a guarantee) save a few weeks of CapEx costs compared to using an optimized solution like Kafka. However, 6-12 months later, you may find yourself back to square 1 when the Postgres extension fails to scale. At that point, migrating to a purpose-built technology becomes an arduous task-your system has grown, and now it may take months of effort and a larger team to make the switch. Ultimately, this approach can cost more time/money than starting with a purpose-built solution from the start. I’ve seen this firsthand at Citus, where customers like Cloudflare, Heap etc eventually moved to dbs like ClickHouse, SingleStore etc. While these migrations happened a few yrs later, times have changed - data grows faster now, and the need for a purpose-built dbs arises much sooner. It’s also worth noting that Citus was an incredible piece of tech that required years of dev work. tl;dr-think carefully before choosing the right tech as you scale. Cramming everything into Postgres might not be the best approach for scaling your business.

    KronisLV 4 days ago | root | parent | prev | next [–]

    KronisLV 4 days ago | root | parent | prev | next [–]

    news.ycombinator.com

Similar pages

Browse jobs

Funding

PeerDB (YC S23) 2 total rounds

Last Round

Seed

US$ 3.6M

See more info on crunchbase