AI help is now live in the #questdb docs.
Ask questions, get instant, context-aware answers on SQL, ingestion, and performance.
No fluff—just clear support for your workflow.

#AI #devtools

LLMs + REST APIs = powerful combos.

Here’s Claude Code querying a ~1BN row crypto dataset from QuestDB using plain SQL over HTTP. No SDKs, just curl.

Clean pattern for real-time insights with minimal setup. Screenshot below.
#QuestDB #LLM #SQL #RESTAPI #TimeSeries

We’ve spent months making QuestDB 9.0 the fastest, most robust time-series DB we’ve ever shipped.

✔️ N-dimensional arrays
✔️ Smarter materialized views
✔️ Ultra-low latency
✔️ Better UX

Built with your feedback

questdb.com/blog/questdb...

#questdb #githubrelease #opensource

Kicking off a new series on order book analytics in SQL with #questdb: bid-ask spread, liquidity thresholds, imbalance, quote stuffing—directly over L2 snapshots. Follow along for real SQL examples this week.

#timeseries #trading #orderbook #FX #SQL

Real-time analytics with an all-in-one system: Are we there yet? | QuestDB Explore how modern databases are evolving into unified real-time analytics platforms. Compare TimescaleDB, ClickHouse, InfluxDB, and QuestDB's approaches to handling both historical and streaming…

Real-time analytics is still too complex—too many tools, too much glue.

Are we close to a true all-in-one solution?

#QuestDB 's Marko Topolnik explores how materialized views simplify things + compares QuestDB, ClickHouse, TimescaleDB & InfluxDB.
questdb.com/blog/realtim...

Middle of a dashboard refresh is a bad time for GC.

QuestDB sidesteps it with:

* Off-heap memory
* Memory-mapped partitions
* Native C++/Rust in the hot path

What you need. Where you need it. When you need it.

questdb.com/docs/guides/...

We watched Interstellar. That movie is a trip!

Slow dashboards make us feel like the wave planet.

We like 'em live, like our fresh crypto dashboards.

👉 questdb.com/dashboards/c...

Somewhere in a database... a lonely little row from 2017 still holds out for hope.

QuestDB’s TTL helps old data move on.

Listen, row, don't think of it as deletion... More like closure.

We just need some time for ourselves (and our more recent data).

questdb.com/docs/concept...

This is your database when you don’t use materialized views to pre-compute aggregates.

Every dashboard refresh is like a hot iron to the foot.

Save your DB('s feet)! Materialize your views.

questdb.com/blog/materia...

Live Crypto Price Charts & Market Data | QuestDB See QuestDB queries in action on a live, real-time cryptocurrency analytics dashboard, powered by Grafana.

Our real-time crypto dashboards received a fresh coat of paint.

New dashboards featuring:

* Candles
* Coefficients
* VWAPs
* More

Ooh-hoo, yep 'lotta neat stuff. And it's all fast and fluid.

To the moon, and such!

👉 questdb.com/dashboards/c...

MCP is shiny and new.

Vector databases are shiny and new, too.

What can a vector DB do that a high-performance DB + frontier model via MCP can’t?

Genuine inquiry! — Please share your use case.

A common typo: WAL log (Write-Ahead Log log)

Or is it?!

Write-Ahead Logs (WAL) actually do log twice.

First to the WAL.

Then to the DB.

The more you know! 🌈

New in QuestDB docs: full breakdown of our architecture.
Learn how we achieve high-performance on time-series data — from storage and ingestion to the query engine.

Built from scratch in low-latency Java, C++ & Rust.

📘 Explore it here: questdb.com/docs/guides/...

Live crypto charts, tick-by-tick 🟢
OKX ➜ QuestDB ➜ Grafana (⏱ ms latency)

• Trade blotter & order-book depth
• OHLC / VWAP / RSI / Bollinger Bands
• BTC, ETH + 20 pairs

Dive in: questdb.com/dashboards/c...

How to create a materialized view | QuestDB Learn what a materialized view is, its benefits and drawbacks, and how to quickly create materialized views in QuestDB to supercharge your aggregation queries with incremental refresh.

Tired of heavy SAMPLE BY queries slowing down dashboards in QuestDB?

Use materialized views to precompute OHLC & other aggregates. They auto-refresh on ingest for blazing-fast reads! ⚡

Learn how 👉 questdb.com/blog/how-to-...

#SQL #QuestDB

Why AI needs a database | QuestDB AI won't replace databases — but it will transform how we use them. This article explores why AI and databases work better together, breaking down LLM tokenization, real-time data access, and…

Asking an LLM for an exact price point is like asking a poet for your bank balance.

You might get a moving response…

But it won’t be right.

“ETH price on March 8, 2022?”

“Twas fair and volatile, sir.”

Hmm... You'll need a database.

questdb.com/blog/why-ai-...

QuestDB Release Notes Latest updates and improvements to QuestDB Open Source and Enterprise editions

Sometimes we make cool things and forget to tell people.

Ahem: #QuestDB has a fresh and funky release notes application.

Want to know what's been happening over-time, for open source and Enterprise?

Checkout questdb.com/release-notes/ and explore.

QuestDB now supports Oracle-style comments, so you can add helpful annotations…

Or sneak messages to your coworkers. 👀 ☕️

Why AI needs a database | QuestDB AI won't replace databases — but it will transform how we use them. This article explores why AI and databases work better together, breaking down LLM tokenization, real-time data access, and…

Structured data is very sneaky!

Just ask your favourite LLM.

1GB of text = ~250M tokens
1GB of structured trades = ~1B tokens

That’s 4× the token cost for less semantic richness.

Our new agentic friends have many capabilities, chewing structured data not among them.

questdb.com/blog/why-ai-...

Hammer! Shooting Star! Bullish Engulfing! Piercing Line! While these sound like gnarly video game moves, they are in fact common patterns in OHLC candlestick charts. Learn more about them from our glossary: questdb.com/glossary/ohc... -- and checkout #QuestDB if your charts need a speed boost.

Why AI needs a database | QuestDB AI won't replace databases — but it will transform how we use them. This article explores why AI and databases work better together, breaking down LLM tokenization, real-time data access, and…

Agents, they're suddenly everywhere. They sure seem to bring lot to the table. But there's a weakness. Structured data. Why? Read! questdb.com/blog/why-ai-...

The power of community and openness —customers submitting Pull Requests to the #QuestDB Rust client: github.com/questdb/c-qu....

Pro-tip: it is now possible to change SYMBOL capacity via SQL: questdb.com/docs/referen...

#questdb #symbol #cardinality

Getting from zero to ingesting, querying, and exporting data to Parquet with QuestDB in under 3 minutes. That’s what we showed live at STAC in London—no magic, just fast, SQL-native performance. #questdb #opensource #SQL #Parquet

One Trading - Where smart money does more - Blog - One Trading Partners with QuestDB to Power Low-Latency Market Data Infrastructure The only place you can trade EU regulated crypto futures. Bring your trading ideas to life - buy and sell quickly and easily 24/7.

With QuestDB powering real-time monitoring and historical analysis, their ultra-low latency architecture runs at sub-200μs round-trip ⚡

More in the press release 👇
www.onetrading.com/blog/one-tra...

#QuestDB #OneTrading #TradingTechnology #TimeSeries #Database

Building K-line (Candlestick) Charts with QuestDB and Grafana | QuestDB In this tutorial, we will stream real-time crypto data from polygon.io and utilize QuestDB’s new materialized view functionality to create aggregated OHLC tables efficiently. We will then visualize th...

One of the use cases materialized views unlock is faster Klines —candlestick— charts. In our latest blog post we show an example of ingesting market data from Polygon.io, creating cascading materialized views, then using Grafana.com to display a real-time dashboard.

questdb.com/blog/kline-c...

This is what happens when you put together QuestDB and github.com/finos/perspe..., a data visualization tool designed for large and streaming data. Look at this cool blotter updating every 50ms with market data coming from QuestDB.

Screenshot from slack, showing Jaromir saying: That's excellent news :tada: I included a PHP test runner in the fix. So from now on PHP PDO has the same test coverage as we have for Python, Rust or C#. This should prevent future regressions. Many thanks for your help: The report, reproducer and now the confirmation - it all helped greatly.

And that, kids, is how PHP made it into our codebase.

A user on PHP hit a QuestDB 8.2.2 issue. We gave an immediate workaround, then @jerrinot.bsky.social shipped a permanent fix 11 days later—with a new PHP test runner for coverage on par with Python or Rust. Community collaboration FTW!

We finally benchmarked InfluxDB 3 OSS Core (Alpha) | QuestDB After many years of waiting, we benchmarked InfluxDB 3 OSS Core (Alpha). How does it fare? What are its strengths and weaknesses? We take an initial look, benchmark ingestion, and spot some caveats..

Our team was curious about the performance of the brand new InfluxDB 3 Core (OSS), so we went and tried out their Alpha. We wrote about our findings at questdb.com/blog/influxd...

Airflow web console view showing a graph with four tasks: Convert_to_parquet (postgresOperator), detach_partition (postgresOperator), upload_to_s3 (python_operator), delete_folder (python_operator)

Dagster web UI view showing a graph with five operations: get_execution_date, convert_to_parquet, detach_partition, upload_files_to_s3, and delete_folder

Did you know you can orchestrate and schedule your QuestDB data workflows using Airflow with the PostgresOperator or Dagster with a resource using psycopg?