🎩 of DevRel & Developer 🥑 @elastic Tweets about Elasticsearch, Kibana, search, observability, security — optimized for relevance DMs are open
San Francisco
- Tweets 22,901
- Following 736
- Followers 5,212
- Likes 12,163
Pinned Tweet
6,000,000,000 downloads of elasticsearch, kibana, logstash, beats, elastic agent
to everyone who downloaded it, built on it, contributed to it, deployed it, broke it, fixed it, and made something amazing with it: thank you
ALT 6B downloads of the ELK
1
3
14
567
hugops to all the PHP users — php2shell, an unauthenticated PHP core RCE
looking at what the exploits do and how to detect them: elastic.co/security-labs/thr…
3
3
300
deja vu. and funny how we've gone from "we need to upgrade faster" to "we should only upgrade with an enforced delay" 😬
🚨 LiteLLM, a Python package for connecting to AI model providers with 92M+ monthly downloads, has been compromised.
Version 1.107.0 ships credential-stealing malware through PyPI wheels and Docker images, targeting cloud credentials, developer secrets, and AI tools.
1
1
317
I used Opus 5.5 on max and GPT-6 Astra for 5 days straight and I reversed engineered Elastic Security and extracted all the rules. I made them available here:
github.com/elastic/detection…
You’re welcome.
2
273
we’re hanging out at #ObservabilityCon today — join us for a coffee (or something stronger later) and talk observability & your favorite tools
luma.com/RESTlounge
ALT elkie
ALT coffee
ALT more coffee
1
1
15
555
new mural in the @elastic SF office. cool to see it step by step from the blank wall to finish, and find the small details 🪵
ALT final
ALT steps
ALT detail
1
10
329
tired: AI generated slides
wired: mastra's @abhiaiyer live drawing his
time to find more engaging formats when generating stuff is cheap (and boring)
ALT drawing slides
2
2
13
519
Hack the North at UWaterloo — robots, software, and badge hacks with ~1,000 students in 36h. wild progress with LLMs. also scary how much the abstraction layer has shifted; asking about implementation details is like asking about the outcome in compiled code 10 years ago — it exists but isn‘t relevant for most (and we can have opinions on that).
great to bring #elasticsearch to the table (literally) and see how a new generation is picking it up
1
9
279
when I see "Schlüsselwortsuche" for keyword search in a german text...
those are the people who also translate queue ("Schlange"), stack ("Stapel"), or heap ("Haufen")
1
1
277
wild times ahead. we had our fair share of issues around inference as well already — though no deep fakes yet
we got hit with a crazy scam attempt
someone signed up for our inference and did a video call with our sales person confirming intent to spend $400,000 over the next year and asked for increased rate limits
our sales person felt like something was off so he looked up the company - found a bunch of videos of the CEO and it looked and sounded like the same person he talked to
so he figured it was real but we started looking deeper
the email they had been talking to us from was a lookalike domain that was registered 4 days ago. all the traffic they were sending looked bizarre and came from foreign IPs
emailed them for more info to see what they'd say. of course they hit us with the "i'm about to get on a plane, can't do anything right now, please move forward"
messaged the real CEO on instagram and yeah he had no idea what we were talking about, the scammer was deepfaking him
be careful out there, things are getting crazy
1
529
after all the attention for duckdb's acquisition by AWS: one of my favorite things is their "why @duckdb" page — duckdb.org/why_duckdb
explainining their tradeoffs and goals; and how they try to achieve those through technical means
ALT duckdb
6
482
it's all about the recall
Be careful when optimizing around metrics like NDCG@10 because they distract you with agents. Was evaluating rerankers. 10% gap with NCDG@10. Seemed huge. Position bias is so coverage is all you need. Your reranker is now just filtering and helping
recover retrieval failures.
1
3
490
"Float Bloat: vector serialization gone wrong" by storing float64 instead of float32. it's a surprising problem but with elasticsearch you're in luck:
* since 9.2 (index creation) the setting `index.mapping.exclude_source_vectors` is on by default — you don't store the floats in source at all any more
* `semantic_text generates` the embeddings server-side and has always avoided the problem
and for even more optimized ingestion, use base64 encoded strings: elastic.co/search-labs/blog/…
full article with examples how not to do it in various programming languages: bonsai.io/blog/float-bloat/
ALT float bloat
2
1
2
427
all those posts floating around with "popular databases and their minimum RAM requirements" with almost random values — probably whatever an LLM said
but the missing question is: for what?
* to start the process?
* local development with a little data?
* production?
1
10
517
hacking a tablet that was EOLed by amazon: "American frontier models won’t help and Chinese will, but not without reasoning about whether they should"
the barrier is to pick the right model, say "keep going" over and over, and paying the token costs: ericpardee.github.io/fire-hd…
2
1
6
456
hacktoberfest is back: hacktoberfest.com
but not as open source contributions but instead as a series of events; maybe not too surprising now that MLH is the main driver
sad to see for open source in general. but probably the saner version for maintainers #sloptoberfest
1
3
242
first hack night in the new @elastic san francisco office together with @mastra tomorrow: build and compete on prices with agent memory
already at 250 RSVP but thanks to the new space we can squeeze in a few more: luma.com/mastra-hack
🤞 for tomorrow
ALT video wall
2
4
405