Real-time distributed OLAP datastore. Point at Kafka and start querying. Join our growing community at https://xitter.cf/t.co/s0Tg6jKf8W
Mountain View, CA
Joined June 2015
- Tweets806
- Following120
- Followers5.1K
- Likes760
Apache Pinot retweeted
Recorded a fantastic episode with Ankit Sultana, PMC at @ApachePinot and SWE at Uber.
We discussed Pinot’s architecture, integration with Kafka, Replication, Segment structures, query engines and much more.
Good news: we also captured on excalidraw.
Stay tuned! Releasing next week.
Subscribe here and support deep technical systems discussions: youtube.com/@thegeeknarrator
Apache Pinot retweeted
Excited to share a key Grocery use-case we had onboarded to @ApachePinot. It powers internal and external user facing tools while handling 100k+ updates per second across billions of rows.
Outside of the usual analytics workload (aggregations, etc.), with the Lucene powered Text Index, we were also able to enable user-facing search use-cases.
Discover how @Uber used Apache Pinot™ to build a real-time index on its massive inventory of billions of items to power search use cases, internal tools, and operational workflows.
read more: uber.com/blog/blazing-fast-o…
#UberEngineering #UberEng #ApachePinot
Apache Pinot retweeted
Did you know there is an @ApachePinot 🍷 connector?
It takes about 5 minutes to connect to, then you can start building dashboards and exploring your data however you want.
buff.ly/mZOQL9r
Apache Pinot retweeted
How did LinkedIn use a sketch algorithm in Apache Pinot to achieve:
• an 88% reduction of data (1TB → 120GB) 🔥
• improve data freshness 50%?
• 5x lower p95 latency?
🧵 In this thread, you'll:
• learn how to compute audience intersection sizes
• see some good memes 😁
Apache Pinot retweeted
Try joining the Apache Pinot Slack community; there are experienced folks who can help ! slack.pinot.live
Apache Pinot retweeted
A Year in Review: @ApachePinot in 2024 🍷
Join us tomorrow as we reflect on an incredible year of innovation and collaboration in the Apache Pinot community! 🎉
linkedin.com/events/apachepi…
Apache Pinot retweeted
Welcome to Apache Pinot 201! Your video curriculum on the advanced topics of @ApachePinot 📹
Join @gAmUssA, our Head of Developer Advocacy, for this 11-part series that will enhance your skills and empower you with practical, actionable insights: youtube.com/watch?v=svNBVY9F…
Apache Pinot retweeted
.@Citi uses @ApachePinot to measure and manage risk. They were able to get data freshness from once *per day* to as fast as 1 second. Billions of events per day on over 1 million positions.
#RTASummit – at San Jose, CA
Apache Pinot retweeted
.@Cisco @Webex uses @ApachePinot. Why? Because it outperformed what they had been using by up to 150×. #RTASummit
startree.ai/resources/cisco-… – at San Jose, CA
Apache Pinot retweeted
.@uber saved $2 million on crash analytics using @ApachePinot. 11,000 new code and infrastructure code changes weekly. #RTASummit
startree.ai/blog/uber-app-cr… – at San Jose, CA
Apache Pinot retweeted
.@sovrnholdings saw their adtech data freshness drop from 25-48 hours down to 0 hours. Subsecond latencies!
startree.ai/resources/sovrn-… – at San Jose, CA
Apache Pinot retweeted
For @startreedata it means using @ApachePinot. Real-time upserts are needed to update rows of data.
You have to be operating on the freshest of data. It has to be indexed the right way.
And it needs to be *affordable.*
#RTASummit – at San Jose, CA
Apache Pinot retweeted
Now, you have go support hybrid tables in @ApachePinot — Lambda architecture — to query both real-time and offline [batch] data sources.
#RTASummit – at San Jose, CA
Apache Pinot retweeted
Also, because a lot of data is upserted, it makes more sense to do query-time JOINs, rather than doing pre-JOINs. This uses the multi-stage query engine. Read more here how they work: startree.ai/blog/query-time-…
#RTASummit #web3 @magiceden @startreedata @ApachePinot – at San Jose, CA
Apache Pinot retweeted
Inverted indexes are great, but for pre-aggregation, the star-tree index of @ApachePinot is unbeatable in terms of performance. @MagicEden #Web3 @startreedata #RTASummit
Learn more in this blog:
startree.ai/blog/star-tree-i… – at San Jose, CA
Apache Pinot retweeted
What was @MagicEden using before @ApachePinot? @ApacheSpark. But it required a lot of ETL and manual jobs. Cost was fairly high. #RTASummit – at San Jose, CA
Apache Pinot retweeted
Now they can combine batch data loading from @ApacheSpark with real-time data from @apachekafka into @ApachePinot. #RTASummit #web3 – at San Jose, CA
Apache Pinot retweeted
How big of a difference did it make?
@MagicEden moved from 5 minute [300s] freshness to *10 seconds*, and as fast as <5 seconds. [30× to 60× better] Query latency dropped to *100ms*. Shrinking costs from $30k/mo to $2k/mo — 15× cost savings. #RTASummit @ApachePinot – at San Jose, CA
