
VP of Revenue and Product at @TigerDatabase / @TimescaleDB Formerly VP Product and GM @newrelic.
Barcelona
Joined March 2007
- Tweets1.1K
- Following360
- Followers403
- Likes722
Ramon Guiu retweeted
Great technical deep dive into @TimescaleDB's new Adaptive Integer Compression (AIC).
Compression effectiveness depends heavily on the distribution and entropy of the data. There's no single encoding that works best everywhere, so AIC adaptively chooses among six different encodings for each block.
Combined with SIMD bitpacking and other low-level optimizations, the result is 5–7x faster decompression, 1.5x faster median queries, and ~10% smaller databases.
One particularly fun result: compressed batches small enough to fit inline rather than behind a Postgres TOAST pointer, eliminating thousands of unnecessary fetches.
tigerdata.com/blog/adaptive-…
Ramon Guiu retweeted
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.
⚡️⚡️⚡️
Ramon Guiu retweeted
23 out of 23 system integrators in @WalkerDReynolds' Industry 4.0 Growth Program are using @TimescaleDB.
A pretty remarkable observation from Walker on a recent 4.0 Solutions podcast about how the industrial historian market is changing:
"If you look at our industry ... it really is staggering to me. The industry 4.0 community is over 100,000 people. All of them are using time-series in some way. All of them used to be buying historians. And now this is crazy: none of them are buying historians.
If you look at the 23 integrators we have in the industry 4.0 growth program: all of them are using TimescaleDB. Every single one, their defacto is TimescaleDB.
10 years ago everybody bought a historian. Then in the last 5-7 years, everybody defaulted to, 'I'm going to build my historian using TimescaleDB as the backend.'"
It's exciting to see how much the modern industrial data stack is changing.
I'm attending #AWSSummit in Hamburg next week. Ping me if you'll be there. I would love to chat about Postgres, real-time analytics and time-series.
TimescaleDB will be at #AWSSummit Hamburg.
Let’s talk PostgreSQL, real-time analytics, time-series workloads on AWS, and AI applications built on operational data.
Book a meeting with our team at the event!
tsdb.co/i4nemls3
Ramon Guiu retweeted
Siemens + TimescaleDB: Tested at CERN, built for the world.
The ATLAS Detector is one of the most complex machines ever built, designed to probe the subatomic structure of the universe. Unsurprisingly, the control systems behind it are just as complex.
To support this scientific mission, @CERN researchers turned to @TimescaleDB as the backend for their SCADA platform, built on @Siemens ETM's WinCC Open Architecture.
Today, we're announcing that this native integration will ship to all WinCC OA users, with TimescaleDB becoming the default storage layer for WinCC OA v3.21 later this year.
Excited to be partnering with Siemens ETM here. And another step toward modernizing the data infrastructure for the physical world.
Photo from @hannover_messe yesterday. Announcement below.
🏭🏗️⛏️🛢️⚡️🚚⚙️🦾
Ramon Guiu retweeted
Two big announcements today kicking off @hannover_messe, but one story:
The industrial historian is being modernized. Legacy products are stuck in 1990s technology, and industrial orgs need to do far more with their operational data: more data, faster decisions, and now AI.
Today we announced @TimescaleDB Enterprise: commercial packaging of our Postgres-based time-series database -- HA, backups, monitoring, admin console, optional cloud sync -- for on-prem and edge.
We also unveiled a strategic alliance with @InductiveAuto. End-to-end: Ignition at the edge and plant → TimescaleDB Enterprise at each site → Tiger Cloud across the enterprise.
Physical AI will run on factories, grids, and fleets, and it needs an open, SQL-native time-series foundation from sensor to cloud.
That's the foundation we already power across 1000s of companies.
🏭🏗️⛏️🛢️⚡🚚⚙️🦾
Ramon Guiu retweeted
Our team is heading to @hannover_messe 2026!
Find us in Hall 14, Booth L18 (April 20–24). If you're working with Postgres, time-series, or industrial data, come say hi 👋
Book a meeting: ow.ly/nncx50YJ2gl
Ramon Guiu retweeted
pg_textsearch v1.0 is now GA, on @TigerDatabase Cloud and open source.
A full BM25 search engine -- tokenization, indexing, compression, and query execution -- built in C directly inside Postgres. No sidecar or separate system.
Indexes live in Postgres pages, use the buffer cache, and participate in WAL, replication, and backups. Block-Max WAND for fast top-k, SIMD-accelerated posting lists, and parallel builds (138M docs in <18 min).
On MS-MARCO: 2.4–6.5x faster than ParadeDB/Tantivy (2–4 terms), with 8.7x higher concurrent throughput.
Full architecture and benchmarks:
tsdb.co/5p5sh9i7
Ramon Guiu retweeted
We shipped a lot in Q1.
TimescaleDB 2.25 with up to 289x faster queries on compressed data. Postgres 18 by default. Full-text search that rivals the fastest Postgres search engines. Azure Marketplace self-serve. New AWS region. Better Console UX.
Full roundup: tsdb.co/hdeb5yc2
Start a free trial: console.cloud.timescale.com/…
Ramon Guiu retweeted
New post by @SoPolliog
I Benchmarked TimescaleDB vs ClickHouse vs MongoDB for Observability Data - The Results Surprised Me
dev.to/aws-builders/i-benchm…
#aws #database #opensource #performance
Ramon Guiu retweeted
Fun to watch this in real time: Fortune 500 company migrating an on-prem @TimescaleDB deployment to Tiger Cloud at ~2 Gbps using our managed LiveSync tooling.
Replying to @GlookoInc
@GlookoInc supports 1M+ people living with diabetes and processes 3B+ data points every month to improve care.
They’ve accumulated 75+ billions of health data points — all tied to real patients making daily decisions.
All on Postgres — no complex multi-database stack.
Really proud of what our team at Tiger Data is helping power.
👇
1
1
36
They migrated from a document database. Their results:
• 480x faster queries: from 4 minutes to 500 ms
• 95% compression
• ~40% lower costs
• 100M+ readings inserted per day
You can scale massively on Postgres — and power real healthcare impact.
Full story:
tigerdata.com/blog/how-glook…
10
Ramon Guiu retweeted
🚨 @TigerDatabase (Yes! the team behind TimescaleDB) is quietly revolutionizing how we build agents.
The Old Way: Glueing together Postgres + Vector DBs + Search tools = complexity.
The New Way: Tiger Data’s Agentic Postgres handles it all natively 🔥
What’s now possible 🧵 ↓
✦ Hybrid retrieval → combine keyword filtering and vector similarity in a single query
✦ Forkable databases → retrieval and safe experimentation without external systems
✦ Native Postgres semantics → everything stays queryable, inspectable, and debuggable in SQL
Why this matters:
→ Fewer moving parts in RAG pipelines
→ No extra infrastructure to stitch together
→ Agents can iterate and learn safely
→ Production data stays isolated and protected
This is an awesome shift!
Stop bolting workflows onto your database. Let your database be the workflow 🔥
Ramon Guiu retweeted
LaLiga has been ordering Spanish ISPs to block ~3000 IP addresses almost every weekend. Because Cloudflare IPs are shared, this has been doing massive collateral damage to thousands of legitimate websites, apps, and vital services - all at the whim of a private corporation.
Ramon Guiu retweeted
Big moment for Postgres!
Search has always been Postgres' weak spot, and everyone just accepted it.
If you needed a real relevance-ranked keyword search, the default answer was to spin up Elasticsearch or add Algolia and deal with the data sync headaches forever.
The problem isn't that Postgres can't do text search. It can.
But the built-in `ts_rank` function uses a basic term frequency algorithm that doesn't come close to what modern search engines deliver.
So teams end up:
- Running a separate Elasticsearch cluster just for search
- Building sync pipelines that inevitably drift out of consistency
- Paying for managed search services that charge per query
- Accepting mediocre search relevance because "good enough" ships faster
But this is actually a solvable problem.
You can realistically bring industry-standard search ranking directly into Postgres, which eliminates the need for external infra entirely.
This exact solution is now available with the newly open-sourced pg_textsearch by @TigerDatabase, a Postgres extension that brings true BM25 relevance ranking into the database.
BM25 is the algorithm behind Elasticsearch, Lucene, and most modern search engines. Now it runs natively in Postgres.
Here's what pg_textsearch enables:
- True BM25 ranking with configurable parameters (the same algorithm powering production search systems)
- Simple SQL syntax: `ORDER BY content <@> 'search terms'`
- Works with Postgres text search configurations for multiple languages
- Pairs naturally with pgvector for hybrid keyword + semantic search
That last point matters a lot for RAG apps. The video below shows this in action, and I worked with the team to put this together.
You can now do hybrid retrieval (combining keyword matching with vector similarity) in a single database, without stitching together multiple systems.
The syntax is clean enough that you can add relevance-ranked search to existing queries in minutes.
pg_textsearch is fully open-source under the PostgreSQL license.
You can find a link to their GitHub repo in the next tweet.
Ramon Guiu retweeted
ColumnarIndexScan is the payoff: 289x faster queries on compressed data.
It answers min/max/first/last directly from sparse index metadata, so the scan becomes "read index summaries," not "decompress batches."
By the numbers: 1.1ms with ColumnarIndexScan vs 312ms with ColumnarScan for GROUP BY device + first/last/min/max.
It shines when your GROUP BY and filters line up with the compression layout, so the planner can push the work down into the compressed scan.
Also likely landing in @TimescaleDB v2.25 this January. Postgres, supercharged.
⚡⚡⚡
New @TimescaleDB planner win from Sven Klemm (~v2.25): 50x faster columnar queries.
When a WHERE clause on the primary time dimension is guaranteed true for a partition/chunk, we stop pushing that qual down into the partition plan.
For columnar partitions this is huge: you no longer decompress the time column just to prove the partition is in range.
Result: 50x faster SELECT count(*) … WHERE time > '2025-01-02', and it unblocks ColumnarIndexScan / SkipScan when filters touch compressed cols.
🚀🚀🚀
Ramon Guiu retweeted
Replit's deep dive into their snapshotting infrastructure, posted Thursday, is worth a read. It highlights a key shift: once agents operate on real application state, experimentation becomes the norm.
Agents branch state, explore in parallel, and backtrack frequently. Assumptions that break traditional databases and storage systems.
What struck me is how closely @Replit's architecture mirrors what we built with Fluid Storage at Tiger Data. Different implementations, same conclusion: databases need to be forkable, and that capability has to live in the storage layer.
I wrote up a comparison on what this convergence means for agentic experimentation 👇
Ramon Guiu retweeted
𝗔𝗻𝗼𝘁𝗵𝗲𝗿 𝗛𝗼𝘁 𝗝𝗼𝗯 𝗔𝗹𝗲𝗿𝘁! 🚨
We’re hiring a Senior Developer Advocate (AI / Agentic Postgres) in SF / Bay Area!
If you live and breathe AI dev tools like Cursor, Claude Code, Windsurf, Copilot, and you build with agents every day, this one’s for you. We’re looking for someone ready to define how developers actually build with agents on real infrastructure, not just talk about it.
You’ll work directly with me to shape Agentic Postgres, our next generation database platform for agentic workloads built on PostgreSQL.
This is community-centric, product-influencing work:
🔥 Ship daily demos showing what’s possible with agents
🔥 Co-host SF events like Agents & Coffee and represent us globally
🔥 Create technical content that teach and inspire builders
🔥 Influence both product direction and developer perception
🔥 Build the voice and momentum for Agents as the new developers worldwide
If you’re a builder, teacher, community-spark, and you want to help define the future of databases for AI workflows... let’s talk.
Come help us put Agentic Postgres on every developer’s toolkit.
tigerdata.com/careers/52906f…
Ramon Guiu retweeted
This graph is WILD
github.com/timescale/pg_text…
Ramon Guiu retweeted
We just open-sourced pg_textsearch (permissive Postgres license).
It's our Postgres extension that brings true BM25 ranking directly into Postgres.
If you care about fast, relevance-ranked keyword search without leaving Postgres -- or hybrid retrieval by combining pg_textsearch with pgvector/pgvectorscale -- this is for you.
Go 🌟 it on GitHub!
















