First, make Postgres really good at search. Then make search really easy to use. That's the thinking behind pg_textsearch, pg_vectorsearch, and now searchgres.js.

Really enjoyed joining Colby Clegg, CEO of @InductiveAuto, on the Ignited podcast. I've long admired what they've built with Ignition: an open, flexible platform that has transformed how industrial teams build and deploy software. We talked about our growing partnership, why @TimescaleDB is such a natural fit for industrial workloads, and how we're bringing the two together with the new Ignition Enterprise Historian. Excited about what this unlocks for industrial engineers! youtube.com/watch?v=di0ddOuC…

GitHub source code (Postgres Licensed): github.com/timescale/pg_vect…

Blog post: tigerdata.com/blog/announcin…

Today we're open-sourcing pg_vectorsearch, a new vector search extension for PostgreSQL. The performance improvements over HNSW are pretty remarkable. At 100 million vectors: ◆︎ Faster queries, 4× less RAM: 6 ms vs. 8 ms at 95% recall, using 128 GB vs. 512 GB RAM. HNSW couldn't even build on 128 GB. ◆︎ 7× smaller indexes: 25 GB vs. 180 GB at 100M vectors. ◆︎ 50× faster index builds: 49 minutes vs. 42 hours on identical hardware. And at 1 million vectors, PRISM delivers 15 ms queries at 99% recall on just 2 GB RAM, versus HNSW's 392 ms on 4 GB. What leads to these improvements? Rather than traversing an HNSW graph with random memory accesses, our new PRISM index partitions vectors based on distance, compresses them using RaBitQ, and scans them efficiently using SIMD vectorization techniques. This takes better advantage of modern CPUs and memory locality. This is actually our second open-source vector index. We learned a lot from building DiskANN-based pgvectorscale, including some important limitations of graph-based approaches. Those lessons convinced us there was a better architectural foundation, so we started over. pg_vectorsearch is still in alpha, with more engineering ahead. But we believe fast, scalable vector search should live alongside your relational data, without requiring a separate vector database. Released under the permissive PostgreSQL License, just like our pg_textsearch extension for BM25 keyword search. Together, they provide the building blocks for hybrid vector + keyword search in PostgreSQL. Feedback and contributions welcome! Code and blog post below. ⚡️⚡️⚡️

One of those weeks where you can feel a market shifting in real time. At the Ignition Community Conference, I kept running into existing open-source @TimescaleDB users across data centers, energy, O&G, manufacturing, and other industrial systems, alongside a lot of excitement around the new first-party Ignition Enterprise Historian, powered by TimescaleDB Enterprise. Huge thanks to Colby Clegg, @InductiveAuto's CEO, for inviting me to join him onstage during the keynote to announce our new product offering and strategic partnership in front of 2,000+ attendees. Feels like a real moment for the modern industrial data stack. #ICCUnleashed

Much of the modern world is built on concrete. Apparently, some of that concrete runs on Postgres. :) Titan America uses Tiger Cloud as the data foundation for its manufacturing operations: millions of SCADA sensor readings a day, continuous aggregates for real-time dashboards and business reporting, and 19 TB of time-series data managed across hot and tiered storage. tigerdata.com/case-studies/t…