@EigenVectorizer

@ApacheIggy Core Team | Low latency message streaming @laserdata | OSS | 🎓 @iitmadras '17

Bengaluru
Joined November 2014
That's me. From being an ML Engineer to contributing to a DB project. Never imagined this timeline but here we go.
99 contributors to Limbo! 🤩
3
88
9,743
Krishna Vishal retweeted
it's crazy that Goethe was dropping these bangers in his private journals 200 years before the internet
11
297
13
2,766
43,717
NNT was one of the first people to publicly callout sociologists AFAIK.
Throwback to NNT chacha banger on social science
3
113
Those were the days where I would spend so many hours to get the plots just right, browse through docs, stackoverflow and also ask questions on SO just about matplotlib.
there was a time when you made your last matplotlib figure and you didn't know it was the last one
1
10
605
Krishna Vishal retweeted
One more stream is here. 💚 Today marks the beginning of a new chapter for @laserdatainc We’re bringing together three worlds that have largely evolved independently: 1. Real-time message streaming. 2. Agent Data Fabric. 3. Real-time applications. One infrastructure. One durable foundation. Today we’re announcing major new capabilities across the stack, from VSR-based clustering in @ApacheIggy and Agent Data Exchange (AGDX) Protocol for agent-agent-system data exchange, to new LaserData Cloud plans (launch, scale, performance, and enterprise) and deployment options across AWS and GCP. All these capabilities are in PRIVATE BETA, will be GA on September 18, 2026. And we’re just getting started. We’ll continue shipping throughout the year. But once in a while, we’ll take a moment to bring together the biggest ideas we’ve been working on and share where we’re taking real-time and agent data infrastructure next. Welcome to the next chapter of LaserData. 🚀
2
3
2
7
471
Krishna Vishal retweeted
Something new is streaming. September 9. 💚
3
1
6
150
I've been building deterministic simulation testing suite for Apache Iggy and its starting to pay dividends now. I also have plans to add PBT, fuzzing, mutation based testing for Iggy. We will get there. Our current focus is to deliver v1 of clustering with DST backing its functionality.
Software engineers: build a principled understanding of verification and testing. Fuzzing, property-based testing, deterministic simulation testing, formal verification. The generation half is done. The other half is yours.
15
298
The Man in the arena is always better than the men in the gallery telling him how they could have done better.
A retard in motion is always better than a genius at rest
1
9
321
That escalated quickly.
You don't actually want job. Nobody dreams of labor. What you want is money. And even money is not what you want. You want the things you can buy with money. But what if you could get the things you want without any money? Anyways, have you heard about robbery?
5
857
Krishna Vishal retweeted
There's no reason for software to be slow anymore: danluu.com/perf-opt/
15
63
10
756
53,242
This made my morning.
Me explaining AI to friends
4
449
we hosted an AI salon with @amuldotexe and @lossfunk built around a single question.: what did you believe about AI a year ago that you don't believe anymore? over 400 people applied to working theories. we split the carefully curated teams into five-person pods to go deeper. then everyone came back together and shared what their group had landed on. we closed with an open floor. people brought a problem they were actually stuck on, and the room worked through it together. a few things came up again and again through the afternoon. whether verification matters more than generation now that code writes itself. whether agent evals are quietly becoming the new unit tests. whether judgement has overtaken output as the scarce skill. thank you to lossfunk for the space and to amul, who steered the room. stay tuned for the next one.
1
19
4
72
9,857
Krishna Vishal retweeted
This is like the subprime of databases… But most (still) don’t recover the WAL correctly in at least two ways cf.: 1. “Can Applications Recover from fsync Failures?” 2. “Protocol-Aware Recovery for Consensus-Based Storage” Single node DBs like SQLite can lose committed transactions on storage faults. And distributed protocols (and DBs built on them) like Paxos and Raft can enter split brain and lose the cluster. The papers are worth a slow and careful read.
Yes, that was the fix that most DBMSes applied for fsyncgate But most still don’t use Direct I/O when they replay the WAL on reboot. So they risk externalizing non-durable data as durable (because they still read from the incoherent kernel page cache). cf. “Can Applications Recover from fsync Failure?” which looked into this in 2020, two years after these fixes, and found that PG, SQLite et al were still vulnerable.
4
22
3
141
15,568
Krishna Vishal retweeted
This is exceptional feat for an Indian team at a global level - the difference between the first & the second place is significant
Replying to @frameo_ai
@frameo_ai just ranked #1 in Physion Labs’ independent benchmark of AI filmmaking agents. 88.7 overall - 20 points ahead of the runner-up. The result that interested me most isn’t the ranking. It’s why the gap exists. 🧵
4
7
1
56
2,999
There are good number of well maintained opensource DICOM viewers. I've recently used one of them. Maybe we need to Google first like the old days instead of asking LLMs 😆.
Dentist did a 3D X-ray of my jaw before a root canal and said I wouldn't be able to open the raw data, it needs specialized software. It's 800 DICOM files. Asked Claude Code to make me a viewer. Two prompts later... this is nicer than what he showed me on his screen.
1
184
Krishna Vishal retweeted
Hear me out, I have an idea on how to expand C++'s syntax in a relatively clean way.
C++26 reflection syntax looks abhorrent
21
66
2
1,494
61,626
More people should embrace SIMD. I'm thankful that I got to profile code, write simd, read assembly and go down the optimization rabbit hole of reducing nano seconds relatively early in my career. It completely changed my outlook and made me want to switch my career from ML to systems software.
My SIMD post was on HN and the general hostility towards understanding how computers work and owning your own outcomes is really depressing. A large amount of people seem to simply have the attitude that they hope someone else pays their bills for them. news.ycombinator.com/item?id…
1
24
1,833
Someone once told me if you are so smart why are you competing when there are opportunities with similar payoffs with way less competition. Opened my eyes.
Replying to @Jatin_exe
young people get pumped by competition, but all grown ups know that avoiding competition is how businesses are built.
3
36
2,140
As @amuldotexe bhrata says. Maybe our only job is to slow down time.
When you realise that going slow is the secret to a regulated nervous system.
2
11
1,406