Timescale

🚀 𝗘𝘃𝗲𝗿𝘆 𝗗𝗲𝘃𝗲𝗹𝗼𝗽𝗲𝗿 𝗶𝘀 𝗡𝗼𝘄 𝗮𝗻 𝗔𝗜 𝗗𝗲𝘃𝗲𝗹𝗼𝗽𝗲𝗿 - 𝗜𝗻𝘁𝗿𝗼𝗱𝘂𝗰𝗶𝗻𝗴 𝗽𝗴𝗮𝗶 𝗩𝗲𝗰𝘁𝗼𝗿𝗶𝘇𝗲𝗿 No need for specialized tools or vector databases—pgai Vectorizer lets you create, sync, and manage embeddings with just one SQL command. 🔹 𝗘𝗺𝗯𝗲𝗱𝗱𝗶𝗻𝗴𝘀 𝗶𝗻 𝗣𝗼𝘀𝘁𝗴𝗿𝗲𝘀: Build, store, and sync embeddings alongside your relational data—no extra infrastructure needed. 🔹 𝗥𝗲𝗮𝗹-𝘁𝗶𝗺𝗲 𝘀𝘆𝗻𝗰, 𝗿𝗮𝗽𝗶𝗱 𝘁𝗲𝘀𝘁𝗶𝗻𝗴: Keep embeddings fresh as your data changes. Test models instantly. 🔹 𝗔𝗹𝗹 𝗶𝗻 𝗣𝗼𝘀𝘁𝗴𝗿𝗲𝘀: Everything you need—embeddings, model access, and AI workflows—all with the SQL you already know. #Postgres #pgaiVectorizer #Postgres #Data #AI #SQL #DevTools #AIDevelopment #PostgresExtensions #AIinSQL

pgai.com

I wrote this article comparing TimescaleDB's SkipScan feature to vanilla #PostgreSQL performance for DISTINCT queries (get me the last row for all IDs) while a 200K rows per second ingest was happening. I'm going to be writing more of these smaller performance pieces (sometimes Timescale related, sometimes Postgres related) - I'd love to hear some suggestions from the LI community 🙂

This morning, I had just finished preparing a video code module that introduces Retrieval-Augmented Generation systems to students in my grad-level 𝘐𝘯𝘧𝘰𝘳𝘮𝘢𝘵𝘪𝘰𝘯 𝘙𝘦𝘵𝘳𝘪𝘦𝘷𝘢𝘭 course. Then I saw this post, thanks to a comment by Keegan Reeve. This article includes a lot of helpful observations (and improvements to implementing a database of vector embeddings) for the IR-portion of a RAG system. I hope my students are listening! Thanks, Daniel Ruiz Riquelme for highlighting this post from Timescale!

Many GenAI applications leverage Retrieval Augmented Generation (RAG) systems which, at an enterprise level, rely on vector databases as a key component. The current implementation paradigm of these systems treats embeddings as independent data rather than derived data, creating significant operational and maintenance challenges for engineering teams. To address these challenges, Timescale has introduced the "vectorizer" abstraction through pgai vectorizer, an open-source tool compatible with pgvector that automatically keeps embeddings in sync with their source content. Take a look at the article here: https://lnkd.in/ddCdRNUU #GenAI #AIEngineering #MachineLearning #ArtificialIntelligence #LLM

If there’s some IT stuff that I like more than Event Sourcing then it’s #PostgreSQL. Why? I started as most of the .NET devs mesmerised by MSSQL. Dpn’t get me wrong, it’s a good database, it served me well. But PostgreSQL has superpowers. The biggest is the well-thought engine. It’s built natively to allow extensions. That’s a basics for extending it with tools like Timescale, PostGIS. But even vanilla feature like partitioning, logical replication is impressive on it’s own. Today on Bern .NET group I tried to show that in practice and show on the example building Fleet Management reporting and alerting system joining all of that. Thanks René Leupold for an invitation and all folks that joined me this evening. I hope they had fun 🙂 You can also check PostgreSQL superpowers in my blog: https://lnkd.in/ebnRMwYM

🇧🇷 Hey, Brazil! Join us today (11/07) at PGConf.Brazil at 17:30! Jônatas Paganini will walk us through pgvector, a PostgreSQL extension for storing, indexing, and searching vector data—perfect for machine learning, NLP, and IoT applications.Come and learn about pgvector and the math behind the embedding that makes semantic search work. 🔥 Nos vemos lá! Venha aprender como o pgvector pode transformar suas aplicações com PostgreSQL. 💪 https://lnkd.in/g-CEHy2H

Timescale DB comes with a controversial but TBH sound take on Vector Databases - "they are the wrong abstraction": In production AI applications, managing vector embeddings tends to bring complexities due to vector databases treating embeddings as standalone data disconnected from their source - this is raised as an abstraction that should be addressed as it would otherwise lead to synchronization issues and stale data. TimescaleDB proposes treating embeddings as derived data similar to database indexes, which is interesting given recent extensions from DBs like planetscale to integrate embeddings natively into indexes, similarly through a "native vectorizer" abstraction. In this case however they still leverage the OSS pgai Vectorizer for PostgreSQL which helps automating the synchronization of embeddings with their source data within the database, however this does provide an insight on some of the open challenges that are yet to be addressed, and perhaps also one of the reasons why we still see lack of standardisation on VectorDBs in the State of Prod ML 2024 Survey. Blog: https://lnkd.in/eWk6e_4G -- If you liked this article you can join 60,000 practitioners for weekly tutorials, resources, OSS frameworks, and MLOps events across the machine learning ecosystem: https://lnkd.in/eRBQzVcA #ML #MachineLearning #ArtificialIntelligence #AI #MLOps #AIOps #DataOps #augmentedintelligence #deeplearning #privacy #kubernetes #datascience #python #bigdata

𝙄𝙢𝙥𝙧𝙤𝙫𝙚 𝙌𝙪𝙚𝙧𝙮 𝙋𝙡𝙖𝙣𝙨 𝙗𝙮 "𝘾𝙤𝙣𝙨𝙩𝙞𝙛𝙮𝙞𝙣𝙜" 𝙀𝙭𝙥𝙧𝙚𝙨𝙨𝙞𝙤𝙣𝙨 🔥 🇧🇷 Brazil! We're attending PGConf.Brazil, so come say hi! 👋 Learn how Timescale enhances PostgreSQL query plans through "constification" — a technique to replace function calls with constants during planning. This optimization improves query performance, especially for partitioned tables, by enabling early partition pruning and reducing execution time. 📅 𝘿𝙖𝙩𝙚 & 𝙏𝙞𝙢𝙚: 2024-11-07, 10:30 🗣️ 𝙎𝙥𝙚𝙖𝙠𝙚𝙧: Fabrízio de Royes Mello, Software Engineer at Timescale 🔗 https://lnkd.in/gQC6fdwn

Navigating hierarchical data like graphs and trees data-structures in SQL is tough. In my latest article with Timescale we dive into Recursive Queries with use-cases and example! https://lnkd.in/gmnTGFBU