Abhinav Gyawali

Feldera in 2025: Building the Future of Incremental Compute Feldera's 2025 year in review: comprehensive SQL support, state-of-the-art infrastructure, advanced connectors, and the future of real-time analytics.

2025:
📦 166 unique releases, 1,162 changes, avg. new release every 2.4 days
📊 10x cost reduction for users, hours old insights into sub-second latency
⚡ 70-node Spark clusters -> single digit Feldera instances

2026: Make incremental compute inevitable

Full story: www.feldera.com/blog/feldera...

Done with grad school applications and suddenly I have 5 extra hours in a day

In the past few days, we’ve gone from democracy to fascists killing kids and trying to enforce censorship to absolute wild west anarchy. It is incredibly difficult to be a functioning human being right now.

Most of all it feels like the entire revolution was a play by the same old politicians. 🇳🇵

Also its not just about the social media ban, its against the normalized corruption, theft, total lack of care about the citizens and censorship.

not people but kids*
real bullets*

All animals are equal but some animals are more equal than others. The corrupt Government of Nepal today open fired against kids protesting in school uniform with 14 reported fatalities. This is full blown fascism!!!

We're hiring for two very high-leverage, high-learning roles at Feldera! If you're deeply technical and love seeing customers succeed using a revolutionary compute platform, we'd love to have you on the team 🚀

🚀 We’re hiring an engineer to push Feldera’s automated testing platforms to the limit!

If you get excited about breaking real-world data-intensive systems, chaos testing, fault injection, differential testing and obsessing over performance, this job is for you.

jobs.ashbyhq.com/feldera/709c...

How Feldera Customers Slash Cloud Spend (10x and beyond) By only needing compute resources proportional to the size of the change, instead of the size of the whole dataset, businesses can dramatically slash compute spend for their analytics.

"This is the true power of incremental compute. By only needing compute resources proportional to the size of the change, instead of the size of the whole dataset, businesses can dramatically slash compute spend for their analytics."

👇

www.feldera.com/blog/how-fel...

Good news (for me!) my gender bias paper from 2023 still replicates with GPT-5.
Bad news (for everyone!) my gender bias paper from 2023 still replicates with GPT-5.
arxiv.org/pdf/2308.14921
hkotek.com/blog/gender-...

"why are rust developers so insistent on rewriting tools in rust?"

we're tired of dealing with segfaults in the rest of our tech stack

Interview with Ben Pfaff

Today, the USENIX newsletter ;login: published an interview with me: www.usenix.org/publications.... Read it to find out about Open vSwitch, @feldera.bsky.social, and more!

I *still* keep learning about new pieces of widely used software @benpfaff.bsky.social written. From GNU Autoconf packaging in Debian, to GNU PSPP and to graphics contributions to the Linux Kernel. 🙇

My startup, @feldera.bsky.social, is looking for a remote #rust, Kubernetes, cloud engineer! Our founders are in the US but we have employees in Europe and Asia as well. See jobs.ashbyhq.com/feldera/9754... for details. You can apply there; feel free to post or DM questions. #rustlang

oh so everyone is like "use signal, use tor" and now that the USG is following that advice everyone's all mad?!?!???

The Pain That is Github Actions Developing a complex CI pipeline with github actions is full of gotchas.

with how good the rest of github is, actions could use some love www.feldera.com/blog/the-pai...

From Traces to Insights: How to Analyze OpenTelemetry Data in Real-Time with SQL Analyzing OpenTelemetry data to generate insightful visualizations using Feldera.

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...

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

Incremental Computation over object storage!

Solving Fine-Grained Authorization by Turning the Problem on its Head Build a high-performance policy engine with only a few lines of SQL.

Have you heard of fine-grained authorization (FGA)? It uses a lot of compute, which makes it expensive. But with Feldera, you can implement it more cheaply using incremental computation. See the full story in the @feldera.bsky.social blog at: www.feldera.com/blog/fine-gr...

YouTube video by Feldera, Inc. Feldera Webinar: Incremental Billion Cell Spreadsheets

What a great demo & presentation. Easily one of the most interesting tech demos I've seen in a half decade.

Tired of slow batch processing as your data grows? @felderainc 's incremental processing maintains lightning-fast performance no matter your data size. Perfect for batch, real-time, or hybrid processing needs! ⚡️
www.feldera.com/blog/batch-p...

My mind is currently being blown running the demo locally. SQL gets compiled into Rust, you can define Rust UDFs, etc.

Last time I really did any "managed" stream processing (with something like Spark, Flink) was pre Apple Silicon so feels near an order of magnitude improvement locally. Staggering.

The changes proposed here are quite exciting!

SQL is huge and is implemented differently by different vendors. Ideally, we would have a single specification and consistent implementation, which would make evolving SQL with time easier.

Hello @bsky.app!👋We’re @feldera.bsky.social, your pals in the world of incremental computation! 🚀 Our super fast compute engine let's you incrementally evaluate arbitrarily complex SQL against live and historical data sources.
Smash that follow button and join us on this journey!