Ivan Burmistrov
@isburmistrov.bsky.social
Staff Software Engineer at Meta
Wow, this looks quite great. And this announcement has a link to my post about wide events 🤯
clickhouse.com/blog/clickst...
clickhouse.com/blog/clickst...
ClickStack: A High-Performance OSS Observability Stack on ClickHouse
We're delighted to announce ClickStack: the open-source observability stack built on ClickHouse - logs, metrics, traces, and session replay in one blazing-fast, developer-friendly platform.
clickhouse.com
May 30, 2025 at 8:02 PM
Wow, this looks quite great. And this announcement has a link to my post about wide events 🤯
clickhouse.com/blog/clickst...
clickhouse.com/blog/clickst...
A famous Meta's Scuba UI reimolenented from Scratch with DuckDB as a backend!
github.com/ezyang/scuba...
Scuba is a really awesome internal system at Meta. It's UI is simple but intuitive and powerful, I wrote about it some time ago: isburmistrov.substack.com/p/all-you-ne...
github.com/ezyang/scuba...
Scuba is a really awesome internal system at Meta. It's UI is simple but intuitive and powerful, I wrote about it some time ago: isburmistrov.substack.com/p/all-you-ne...
GitHub - ezyang/scubaduck
Contribute to ezyang/scubaduck development by creating an account on GitHub.
github.com
May 23, 2025 at 6:35 PM
A famous Meta's Scuba UI reimolenented from Scratch with DuckDB as a backend!
github.com/ezyang/scuba...
Scuba is a really awesome internal system at Meta. It's UI is simple but intuitive and powerful, I wrote about it some time ago: isburmistrov.substack.com/p/all-you-ne...
github.com/ezyang/scuba...
Scuba is a really awesome internal system at Meta. It's UI is simple but intuitive and powerful, I wrote about it some time ago: isburmistrov.substack.com/p/all-you-ne...
Reposted by Ivan Burmistrov
My talk on QCon London 2024 has been opened to everyone. I discussed the challenges of building a realtime feature engineering platform.
www.infoq.com/presentation...
www.infoq.com/presentation...
The Harsh Reality of Building a Real-time ML Feature Platform
Ivan Burmistrov shares how ShareChat built their own Real-time Feature Platform serving more than 1 billion features per second, and how they managed to make it cost efficient.
www.infoq.com
March 20, 2025 at 8:52 PM
My talk on QCon London 2024 has been opened to everyone. I discussed the challenges of building a realtime feature engineering platform.
www.infoq.com/presentation...
www.infoq.com/presentation...
My talk on QCon London 2024 has been opened to everyone. I discussed the challenges of building a realtime feature engineering platform.
www.infoq.com/presentation...
www.infoq.com/presentation...
The Harsh Reality of Building a Real-time ML Feature Platform
Ivan Burmistrov shares how ShareChat built their own Real-time Feature Platform serving more than 1 billion features per second, and how they managed to make it cost efficient.
www.infoq.com
March 20, 2025 at 8:52 PM
My talk on QCon London 2024 has been opened to everyone. I discussed the challenges of building a realtime feature engineering platform.
www.infoq.com/presentation...
www.infoq.com/presentation...
March 15, 2025 at 9:24 AM
Reposted by Ivan Burmistrov
Andreas Saudemont, @sarna.dev, and @isburmistrov.bsky.social have a long track record of achieving AI/ML performance at scale. There is still time to join and learn from them at Monster Scale Summit on March 11-12. RSVP for free here > www.scylladb.com/monster-scal...
#ScyllaDB #techtips
#ScyllaDB #techtips
March 10, 2025 at 3:00 PM
Andreas Saudemont, @sarna.dev, and @isburmistrov.bsky.social have a long track record of achieving AI/ML performance at scale. There is still time to join and learn from them at Monster Scale Summit on March 11-12. RSVP for free here > www.scylladb.com/monster-scal...
#ScyllaDB #techtips
#ScyllaDB #techtips
The Scylla team’s conferences—P99 CONF and Monster Scale Summit— are top-tier. The hardest part is choosing a track when every session looks great. And the swag is in its own league :)
Next week’s Monster Scale Summit will be scary good. Keynotes include @kelseyhightower.com, @martin.kleppmann.com, @chris.blue, @gwenshap.bsky.social, @rstephens.me & @adamhjk.me. Given the price of admission (zero) and travel time (also zero), there’s really no excuse to miss this!
March 7, 2025 at 9:23 AM
The Scylla team’s conferences—P99 CONF and Monster Scale Summit— are top-tier. The hardest part is choosing a track when every session looks great. And the swag is in its own league :)
That's actually a pretty good analogy. And bicycle is healthier, btw 😏
By bicycle, it takes about 15 minutes to get from my place to Wimbledon Village.
In a supercar that's 20x as fast, it takes about 15 minutes to get to Wimbledon Village.
In case you were wondering why faster code generation makes little difference to team productivity.
In a supercar that's 20x as fast, it takes about 15 minutes to get to Wimbledon Village.
In case you were wondering why faster code generation makes little difference to team productivity.
February 10, 2025 at 10:45 AM
That's actually a pretty good analogy. And bicycle is healthier, btw 😏
Battle human vs. AI happened today. My 4 y.o. daughter wanted to re-watch a Peppa Pig episode where kids did an animation. I asked ChatGPT and Perplexity, and they both failed. Then I started showing my daughter random Peppa Pig episode titles, and she recognised the right one! Score 1 for humans 🏆
January 2, 2025 at 10:48 AM
Battle human vs. AI happened today. My 4 y.o. daughter wanted to re-watch a Peppa Pig episode where kids did an animation. I asked ChatGPT and Perplexity, and they both failed. Then I started showing my daughter random Peppa Pig episode titles, and she recognised the right one! Score 1 for humans 🏆
Naming matters. Started hearing the phrase "let me plex it" (ask Perplexity). There is no similar sticky alternative with ChatGPT
December 27, 2024 at 2:19 PM
Naming matters. Started hearing the phrase "let me plex it" (ask Perplexity). There is no similar sticky alternative with ChatGPT
Reposted by Ivan Burmistrov
To understand how software engineering will change when anyone (technical or not) can hire an AI agent to spit out code:
How did filmmaking and the movie industry change now that everyone has a professional-camera (previously unattainable to non-pros) in their pocket?
How did filmmaking and the movie industry change now that everyone has a professional-camera (previously unattainable to non-pros) in their pocket?
December 21, 2024 at 6:18 AM
To understand how software engineering will change when anyone (technical or not) can hire an AI agent to spit out code:
How did filmmaking and the movie industry change now that everyone has a professional-camera (previously unattainable to non-pros) in their pocket?
How did filmmaking and the movie industry change now that everyone has a professional-camera (previously unattainable to non-pros) in their pocket?
Opened Twitter. The whole feed is celebrating the achievement of AGI. Good that don't see it here, AGI-free feed feels refreshing.
December 21, 2024 at 12:09 AM
Opened Twitter. The whole feed is celebrating the achievement of AGI. Good that don't see it here, AGI-free feed feels refreshing.
The most beautiful UUIDs will be expensive soon, be quick grabbing one while they're available for free.
how do you all remember every UUID? I find it really hard. so I wrote them all down on every uuid dot com
the list has fast search across all 2^122 values (so you can find your favorites) - hoping to add some social features like "trending UUIDs" soon!
the list has fast search across all 2^122 values (so you can find your favorites) - hoping to add some social features like "trending UUIDs" soon!
December 7, 2024 at 6:23 PM
The most beautiful UUIDs will be expensive soon, be quick grabbing one while they're available for free.
I've enjoyed listening to this podcast with @gradybooch.bsky.social. Grady is formulating his thoughts in a so clear way, and providing relevant historic references in a pretty entertaining way.
newsletter.pragmaticengineer.com/p/software-a...
newsletter.pragmaticengineer.com/p/software-a...
Software architecture with Grady Booch
Today, I’m thrilled to be joined by Grady Booch, a true legend in software development. Grady is the Chief Scientist for Software Engineering at IBM, where he leads groundbreaking research in embodied...
newsletter.pragmaticengineer.com
December 7, 2024 at 4:04 PM
I've enjoyed listening to this podcast with @gradybooch.bsky.social. Grady is formulating his thoughts in a so clear way, and providing relevant historic references in a pretty entertaining way.
newsletter.pragmaticengineer.com/p/software-a...
newsletter.pragmaticengineer.com/p/software-a...
Wow, I just saw a post on LinkedIn where someone debated whether code reviews are needed because, essentially, they don't want to read what AI tools generated. Like, "AI knows better, what's the point". I wanted to comment something, but can't quite find the words.
November 28, 2024 at 4:59 PM
Wow, I just saw a post on LinkedIn where someone debated whether code reviews are needed because, essentially, they don't want to read what AI tools generated. Like, "AI knows better, what's the point". I wanted to comment something, but can't quite find the words.
Read a bit of LinkedIn feed. "Looking for Java developers", "..Rust developers", "...C++ developers",...
I've never associated myself with a single programming language - worked with C#, php (Hack), C++, Python, Java, JS, Go, Kotlin, ... What's the point with limiting the scope to a single one?
I've never associated myself with a single programming language - worked with C#, php (Hack), C++, Python, Java, JS, Go, Kotlin, ... What's the point with limiting the scope to a single one?
November 22, 2024 at 7:45 AM
Read a bit of LinkedIn feed. "Looking for Java developers", "..Rust developers", "...C++ developers",...
I've never associated myself with a single programming language - worked with C#, php (Hack), C++, Python, Java, JS, Go, Kotlin, ... What's the point with limiting the scope to a single one?
I've never associated myself with a single programming language - worked with C#, php (Hack), C++, Python, Java, JS, Go, Kotlin, ... What's the point with limiting the scope to a single one?
Someone needs to make "starter pack" feed with recsys and all of that 😃
HOLY HELL THERE ARE 63000 STARTER PACKS
blueskydirectory.com/starter-pack...
ugh really regret using uint16 in that schema...
blueskydirectory.com/starter-pack...
ugh really regret using uint16 in that schema...
All - Bluesky Directory
A curated collection of all things relating to the Blue Sky social media platform.
blueskydirectory.com
November 20, 2024 at 10:34 AM
Someone needs to make "starter pack" feed with recsys and all of that 😃
Thanks to the quoted post by @mitsuhiko.at, I spent some time refreshing my memory about Project Loom and how it's different from Kotlin Coroutines and whether the latter can benefit from the former. Here is a couple of interesting links:
tinyurl.com/ycybhsuz
tinyurl.com/ykbvzuhz
tinyurl.com/ycybhsuz
tinyurl.com/ykbvzuhz
November 15, 2024 at 7:29 PM
Thanks to the quoted post by @mitsuhiko.at, I spent some time refreshing my memory about Project Loom and how it's different from Kotlin Coroutines and whether the latter can benefit from the former. Here is a couple of interesting links:
tinyurl.com/ycybhsuz
tinyurl.com/ykbvzuhz
tinyurl.com/ycybhsuz
tinyurl.com/ykbvzuhz
Reposted by Ivan Burmistrov
Had a wonderful discussion with @apurvamehta.com and Yingjun Wu on Stream processing.
Also thanks to @alexmillerdb.bsky.social @qianli.dev @isburmistrov.bsky.social @whoisdavid.bsky.social @zychr.bsky.social for wonderful questions on my previous post.
Episode will be released soon! Stay tuned!
Also thanks to @alexmillerdb.bsky.social @qianli.dev @isburmistrov.bsky.social @whoisdavid.bsky.social @zychr.bsky.social for wonderful questions on my previous post.
Episode will be released soon! Stay tuned!
November 13, 2024 at 7:55 PM
Had a wonderful discussion with @apurvamehta.com and Yingjun Wu on Stream processing.
Also thanks to @alexmillerdb.bsky.social @qianli.dev @isburmistrov.bsky.social @whoisdavid.bsky.social @zychr.bsky.social for wonderful questions on my previous post.
Episode will be released soon! Stay tuned!
Also thanks to @alexmillerdb.bsky.social @qianli.dev @isburmistrov.bsky.social @whoisdavid.bsky.social @zychr.bsky.social for wonderful questions on my previous post.
Episode will be released soon! Stay tuned!
Reposted by Ivan Burmistrov
Awesome proposal: region-based memory management.
Bulk deallocation with a small, clean API while falling back to the global GC for corner cases. The design reminds me the ‘deep modules’ from Philosophy of Software Design.
Worth a look: https://buff.ly/4hRzTaO #golang
Bulk deallocation with a small, clean API while falling back to the global GC for corner cases. The design reminds me the ‘deep modules’ from Philosophy of Software Design.
Worth a look: https://buff.ly/4hRzTaO #golang
memory regions · golang go · Discussion #70257
I'm starting this discussion to collect early feedback on a draft design for a kind of region-based memory management in Go. There is no prototype yet, only a design and a preliminary evaluation. P...
buff.ly
November 9, 2024 at 10:12 AM
Awesome proposal: region-based memory management.
Bulk deallocation with a small, clean API while falling back to the global GC for corner cases. The design reminds me the ‘deep modules’ from Philosophy of Software Design.
Worth a look: https://buff.ly/4hRzTaO #golang
Bulk deallocation with a small, clean API while falling back to the global GC for corner cases. The design reminds me the ‘deep modules’ from Philosophy of Software Design.
Worth a look: https://buff.ly/4hRzTaO #golang
www.cncf.io/blog/2024/11...
Good thing in general, although I wish there was OpenTelemetry "core" and "plugins" (tracing, logs, CI/CD,...). Would be easier for vendors to tell which "plugins" they support.
Good thing in general, although I wish there was OpenTelemetry "core" and "plugins" (tracing, logs, CI/CD,...). Would be easier for vendors to tell which "plugins" they support.
OpenTelemetry Is expanding into CI/CD observability
SIG post by Dotan Horovits and Adriel Perkins, Project Leads, SIG CI/CD Observability, OpenTelemetry We’ve been talking about the need for a common “language” for reporting and observing CI/
www.cncf.io
November 4, 2024 at 11:13 PM
www.cncf.io/blog/2024/11...
Good thing in general, although I wish there was OpenTelemetry "core" and "plugins" (tracing, logs, CI/CD,...). Would be easier for vendors to tell which "plugins" they support.
Good thing in general, although I wish there was OpenTelemetry "core" and "plugins" (tracing, logs, CI/CD,...). Would be easier for vendors to tell which "plugins" they support.
Reposted by Ivan Burmistrov
If on macOS do yourself a favor and switch to orbstack.dev to run docker containers. Just works. Better.
OrbStack · Fast, light, simple Docker & Linux
Say goodbye to slow, clunky containers and VMs. The fast, light, and easy way to run containers and Linux. Develop at lightspeed with our Docker Desktop alternative.
orbstack.dev
November 4, 2024 at 6:30 PM
If on macOS do yourself a favor and switch to orbstack.dev to run docker containers. Just works. Better.
Adding to the recent discussion about Staff+ career path - pretty cool comparison to the career of top surgeons.
The interesting part is usually the top surgeons are at teaching hospitals. So they’re scaling impact through education (even narrating to students during a surgery) as well as research on new problems or solutions that will be published and shared with others.
November 4, 2024 at 1:53 PM
Adding to the recent discussion about Staff+ career path - pretty cool comparison to the career of top surgeons.
Reposted by Ivan Burmistrov
The usual clickbait: "blazingly fast.. built in Rust": it's this warm feeling like running C++ code in R or Python ("100x faster!") and that attracts people to learn Rust.
The much greater learnings imo are:
- RAII guided by the compiler
- what a modern toolchain should look like
- how memory works
The much greater learnings imo are:
- RAII guided by the compiler
- what a modern toolchain should look like
- how memory works
November 4, 2024 at 8:27 AM
The usual clickbait: "blazingly fast.. built in Rust": it's this warm feeling like running C++ code in R or Python ("100x faster!") and that attracts people to learn Rust.
The much greater learnings imo are:
- RAII guided by the compiler
- what a modern toolchain should look like
- how memory works
The much greater learnings imo are:
- RAII guided by the compiler
- what a modern toolchain should look like
- how memory works
Video of our #p99conf talk about cloud cost in the context of ML Feature Store: youtu.be/X7u3Q3tOG5U?...
The cost topic is far broader than we managed to cover, tons of things have been left unspoken!
The cost topic is far broader than we managed to cover, tons of things have been left unspoken!
P99 CONF 2024 | Feature Store Evolution Under Cost Constraints by David Malinge & Ivan Burmistrov
YouTube video by ScyllaDB
youtu.be
October 28, 2024 at 3:19 PM
Video of our #p99conf talk about cloud cost in the context of ML Feature Store: youtu.be/X7u3Q3tOG5U?...
The cost topic is far broader than we managed to cover, tons of things have been left unspoken!
The cost topic is far broader than we managed to cover, tons of things have been left unspoken!