Profile avatar
lalithsuresh.bsky.social
CEO @ feldera.com, the incremental compute engine for AI, ML and data teams. Formerly a systems researcher in distributed systems, databases, cloud, OS, PL, and networking. Sci-fi and gaming nerd. lalith.in/research
210 posts 809 followers 137 following
Regular Contributor
Active Commenter

Breaking: surge in global productivity now that Slack is down.

oh cool I’ve been meaning to learn anxiety

Light weight in-memory Postgres instances with PGlite are perfect for unit tests. We recently added a `.clone()` method to make it super easy for have a clean state for each test. Just setup the schema and fixtures once, then clone!

It's gratifying to check out the code from 2 weeks ago and see we've improved performance by around 70% for a user's workload. 🤩

Today I learnt about loser trees, thanks to @benpfaff.bsky.social, who's been leading some pretty insane performance engineering work.

For some reason we all stopped doing meetups for a few years. Let's fix that. See you in Santa Clara on Thursday. www.meetup.com/futureofdata...

Are you OTel Curious? Feldera Engineer Abhinav Gyawali breaks down how to compute on streams of OTel data using Feldera. If you're looking to build your own observability solutions, check out this blog! www.feldera.com/blog/opentel...

It's been an exciting few weeks of deep performance engineering work by the team! Besides the usual tools like perf, we're also liking samply (works for MacOS and Linux).

You can learn more about Feldera's capabilities in our interactive sandbox environment. No commitment, just pure exploration. Take the first step towards faster, more efficient data processing. try.feldera.com

Armored Core 6 PVP tournaments are something else.

Thinking more about what features next-gen query engines will have. I wrote about being able to run anywhere (on-device, edge, cloud, BYOC, and so on). I'm thinking they'll also support both batch and stream processing. (e.g. feldera.com / github.com/linkedin/Hoptimator)

It’s unfortunate that Chromebooks have totally captured the K-12 educational market. Imagine if we gave every kid a notebook running Ubuntu Linux. They’d hate it, we’d hate it, school IT would hate it, but computer literacy would skyrocket

The thing about code generating Rust is that you sometimes end up with gems like Tuple675<A1, A2.... A675> ☠️

Unlike traditional databases that slow down with data growth, our fully automatic incremental computation maintains consistent performance. Ideal for both batch and real-time data processing. Learn more about steady performance as your #data grows! www.feldera.com/blog/batch-p... #datastreaming

Michelin chef: I call this dish "essence of bean sprout" [he pulls out a spray bottle and sprays a light mist in my general direction] Me: wow very interesting Michelin chef: now this one's my personal favorite, I call it "quail egg elevated" [birdsong is played over the speakers] Me: fascinating

I'm incredibly proud of our team. A user pulled thousands of lines of SQL from their data warehouse—an absolute beast of a code-generated query—and Feldera evaluated it incrementally in real-time, delivering the exact same result as the batch query in the warehouse.

Hello! We are organizing our first South Bay Systems Coffee Chat next week. Register here: lu.ma/2f8y3fsj (w/ @alexmillerdb.bsky.social @yingjunwu.bsky.social) We're planning some talks in the upcoming weeks. Please follow the official account @southbaysystems.xyz 🦋 Stay tuned 👀

feldera: The Feldera Incremental Computation Engine ★1025 https://github.com/feldera/feldera

Selling enterprise data infra software is 1% inspiration and 99% fighting with IAM.

We've got some artists here: xls.feldera.io 🌞

www.reddit.com/r/rust/comme...

It's ain't much but it's honest work. github.com/feldera/feld...

I take back the nice things I said about the Cursor agent. If I want random parts of my codebase deleted, I can do that just fine on my own.