Prashanth Rao
@tech_optimist
AI engineer @lancedb | prev @kuzudb. Blogging @ https://thedataquarry.com
You might like
Nice work! More and more folks are getting there, slowly, but surely 🚀
Somehow, I have a feeling that there's folks in my networks who are as excited about this as I am 😏 @joshmo_dev is inventing a new niche at the intersection of Rust + AI 🔥. Please grab his time!
Looking forward to doing my first talk (of many more to come) about @lancedb and also to meeting all the coolest folks working on retrieval, agents and RAG! 👇🏽 Will be doing a *lot* more thinking after such an event, I'm sure 😅
Prashanth Rao @tech_optimist, AI Engineer, is speaking at @ScaleByTheBay next week! Don’t miss his session - Agents Need More Than Vector Search - on Tuesday 11/18. This talk will introduce the principles behind LanceDB and show how its AI-native multimodal lakehouse can power…
Is big data getting TOO big? What's up with @github going down every other week?
Awesome recent demo by Pablo Delgado from @Netflix! Searching over video has never been easier (or more scalable) thanks to @lancedb. The demo shows how to use natural language queries to retrieve relevant video frames from petabytes of video, with <1s latency. 🤯 👇🏽👇🏽👇🏽
Here's a gem for the job market: I for one would love to see Herumb @krypticmouse at places like @modal or @cerebras or any of the other cool places that are building the next generation of AI & ML training/inference infra. When Herumb puts his systems programming hat on, he's a…
I just learned that the one-of-a-kind Herumb Shandilya @krypticmouse, who is soon to graduate from Stanford with Master’s in CS, is on the job market for training infra, inference infra, or other ML engineering roles. Here’s a PSA to y’all: you should hire him.
There are so many interesting research directions that are already being taken in this front, and I'm eager to see how it all pans out over the coming years: - Separating a model's reasoning from its memory (they're currently intertwined) - Selectively "forgetting" from a model's…
What's the latest research on how much baked-in knowledge an LLM needs in order to be useful? If I want a specialist coding model can I trim the size of that model down by stripping out detailed knowledge of human history, geography etc? Can we even do that?
A really great long-context benchmark!
Can LLMs accurately aggregate information over long, information-dense texts? Not yet… We introduce Oolong, a dataset of simple-to-verify information aggregation questions over long inputs. No model achieves >50% accuracy at 128K on Oolong!
When it comes to open source, locally-stored, private memories for your application, @lancedb is great option. One Q I have is this: how is it decided at the application layer when something should be "forgotten"? Worth checking out AVDB by @Shashikant86!👇🏽
Still not sure why big AI labs still rely on the File System, Virtual File System for Agent Memory (e.g Claude Code), Whats issue with using Vector Store like @lancedb as agent memory? It's impressive DiskANN can be game changing for context/memory management. The first project I…
The number of times I've felt this in my lifetime is non-trivial
why are you, someone who doesnt hate themselves, using databases written in java in 2025?
I'm unable to predict what will run out sooner: entertaining posts about embedding-based retrieval on HN, or ideas (and the money involved to fund them) in Silicon Valley
So much confusion around embedding-based retriveal and associated technologies that it will continue to provide entertainment on hackernews for decades
United States Trends
- 1. #FinallyOverIt 5,516 posts
- 2. #TalusLabs N/A
- 3. Summer Walker 17.3K posts
- 4. 5sos 21.8K posts
- 5. Justin Fields 10K posts
- 6. #criticalrolespoilers 4,124 posts
- 7. Jets 68.7K posts
- 8. #GOAI 3,048 posts
- 9. Patriots 151K posts
- 10. Wale 33.1K posts
- 11. Go Girl 25.9K posts
- 12. #zzzSpecialProgram 2,788 posts
- 13. Jalen Johnson 8,915 posts
- 14. 1-800 Heartbreak 1,422 posts
- 15. Domain For Sale 12.7K posts
- 16. Drake Maye 21.3K posts
- 17. Judge 205K posts
- 18. Bihar 249K posts
- 19. Robbed You 4,085 posts
- 20. Disc 2 1,008 posts
You might like
-
Voice of Punjab 🇮🇳
@voiceof_punjab -
Bob Belderbos
@bbelderbos -
Harveen Singh Chadha
@HarveenChadha -
GRAIL
@GrailBio -
Sowmya Vajjala
@adyantalamadhya -
Nouha Dziri
@nouhadziri -
Yanai Elazar
@yanaiela -
Eunsol Choi
@eunsolc -
Lubo Smid
@lubosmid -
Han Guo
@HanGuo97 -
Pei Zhou
@peizNLP -
Yoshitomo Matsubara
@yoshitomo_cs -
Gyuwan Kim
@gyuwankim93 -
Sofie Van Landeghem
@OxyKodit -
Wissam Antoun
@wissam_antoun
Something went wrong.
Something went wrong.