Philipp Krenn
@xeraa.mastodon.social.ap.brid.gy
🎩 of DevRel & Developer 🥑 at Elastic — toots about Elasticsearch, Kibana, search, observability, security
[bridged from https://mastodon.social/@xeraa on the fediverse by https://fed.brid.gy/ ]
[bridged from https://mastodon.social/@xeraa on the fediverse by https://fed.brid.gy/ ]
finally got some billing alerts in @Elastic cloud
less relevant with static setups. much more relevant with #serverless projects — don't judge all the projects running here
full docs in […]
[Original post on mastodon.social]
less relevant with static setups. much more relevant with #serverless projects — don't judge all the projects running here
full docs in […]
[Original post on mastodon.social]
December 16, 2025 at 11:14 PM
finally got some billing alerts in @Elastic cloud
less relevant with static setups. much more relevant with #serverless projects — don't judge all the projects running here
full docs in […]
[Original post on mastodon.social]
less relevant with static setups. much more relevant with #serverless projects — don't judge all the projects running here
full docs in […]
[Original post on mastodon.social]
riding into the christmas season #elkontheshelf
December 13, 2025 at 1:36 AM
riding into the christmas season #elkontheshelf
TIL about pyinfra — ansible but python instead of yaml and a lot faster
I'd be interested in faster but I'm not sure I see the full advantage in python; at least not for the simple examples they show
I'd be interested in faster but I'm not sure I see the full advantage in python; at least not for the simple examples they show
December 11, 2025 at 3:18 AM
TIL about pyinfra — ansible but python instead of yaml and a lot faster
I'd be interested in faster but I'm not sure I see the full advantage in python; at least not for the simple examples they show
I'd be interested in faster but I'm not sure I see the full advantage in python; at least not for the simple examples they show
scoring rubric of #kubecon: content, originality, relevance, speakers, not-a-product-pitch. and no obvious LLM abstracts. sounds widely applicable
https://events.linuxfoundation.org/kubecon-cloudnativecon-europe/program/submission-reviewer-guidelines/#scoring-rubric
https://events.linuxfoundation.org/kubecon-cloudnativecon-europe/program/submission-reviewer-guidelines/#scoring-rubric
December 10, 2025 at 7:34 PM
scoring rubric of #kubecon: content, originality, relevance, speakers, not-a-product-pitch. and no obvious LLM abstracts. sounds widely applicable
https://events.linuxfoundation.org/kubecon-cloudnativecon-europe/program/submission-reviewer-guidelines/#scoring-rubric
https://events.linuxfoundation.org/kubecon-cloudnativecon-europe/program/submission-reviewer-guidelines/#scoring-rubric
fizzy advertising itself as open source — no huge shitstorm yet. maybe it's different for new code-bases
PS: sane choice and a lot of similarities to the elastic license 😅
PS: sane choice and a lot of similarities to the elastic license 😅
December 9, 2025 at 11:26 PM
fizzy advertising itself as open source — no huge shitstorm yet. maybe it's different for new code-bases
PS: sane choice and a lot of similarities to the elastic license 😅
PS: sane choice and a lot of similarities to the elastic license 😅
Agentic AI Foundation (AAIF) bringing MCP, AGENTS.md, and goose to an open foundation with OpenAI, Anthropic, and many more […]
[Original post on mastodon.social]
[Original post on mastodon.social]
December 9, 2025 at 8:03 PM
Agentic AI Foundation (AAIF) bringing MCP, AGENTS.md, and goose to an open foundation with OpenAI, Anthropic, and many more […]
[Original post on mastodon.social]
[Original post on mastodon.social]
side job of the day: making fully monitored coffees at #awsreinvent
making developer dreams come true 😂
PS: dashboards are public on https://ela.st/beans-streams
making developer dreams come true 😂
PS: dashboards are public on https://ela.st/beans-streams
December 3, 2025 at 4:37 PM
side job of the day: making fully monitored coffees at #awsreinvent
making developer dreams come true 😂
PS: dashboards are public on https://ela.st/beans-streams
making developer dreams come true 😂
PS: dashboards are public on https://ela.st/beans-streams
hierarchy of (LLM coding( leverage): how CLAUDE md > specification > research > plan > code
and how to write a good CLAUDE md — but keep it short
https://www.humanlayer.dev/blog/writing-a-good-claude-md
and how to write a good CLAUDE md — but keep it short
https://www.humanlayer.dev/blog/writing-a-good-claude-md
December 3, 2025 at 3:39 AM
hierarchy of (LLM coding( leverage): how CLAUDE md > specification > research > plan > code
and how to write a good CLAUDE md — but keep it short
https://www.humanlayer.dev/blog/writing-a-good-claude-md
and how to write a good CLAUDE md — but keep it short
https://www.humanlayer.dev/blog/writing-a-good-claude-md
But How Do AI Images and Videos Actually Work?
diffusion models, CLIP, and the papers + math behind it — great weekend video: https://www.youtube.com/watch?v=iv-5mZ_9CPY
diffusion models, CLIP, and the papers + math behind it — great weekend video: https://www.youtube.com/watch?v=iv-5mZ_9CPY
November 24, 2025 at 3:10 AM
But How Do AI Images and Videos Actually Work?
diffusion models, CLIP, and the papers + math behind it — great weekend video: https://www.youtube.com/watch?v=iv-5mZ_9CPY
diffusion models, CLIP, and the papers + math behind it — great weekend video: https://www.youtube.com/watch?v=iv-5mZ_9CPY
November 22, 2025 at 1:20 AM
european bureaucracy: it "only" took 2.5 years to get a delayed flight paid out. and after all the costs (including a laywer) you only get 50%. but I'll still take the win 😉
PS: and I have an even older claim still under negotiations 🫠
PS: and I have an even older claim still under negotiations 🫠
November 19, 2025 at 6:55 PM
european bureaucracy: it "only" took 2.5 years to get a delayed flight paid out. and after all the costs (including a laywer) you only get 50%. but I'll still take the win 😉
PS: and I have an even older claim still under negotiations 🫠
PS: and I have an even older claim still under negotiations 🫠
what's the story with langchain vs langgraph?
I thought the focus was switching over to langgraph but the contributions for langchain seem to have taken over again
I thought the focus was switching over to langgraph but the contributions for langchain seem to have taken over again
November 17, 2025 at 4:47 AM
what's the story with langchain vs langgraph?
I thought the focus was switching over to langgraph but the contributions for langchain seem to have taken over again
I thought the focus was switching over to langgraph but the contributions for langchain seem to have taken over again
how to end the week in the office: give elkie a belly rub
November 15, 2025 at 8:17 AM
how to end the week in the office: give elkie a belly rub
if you're in SF on monday, we'll have a hardware oriented meetup together with microsoft (right at the start of #msignite):
* Bad AI In A Box (BAIIAB) — making AI touch grass
* monitor our coffee consumption — I can now live debug coffee from across the world […]
[Original post on mastodon.social]
* Bad AI In A Box (BAIIAB) — making AI touch grass
* monitor our coffee consumption — I can now live debug coffee from across the world […]
[Original post on mastodon.social]
November 14, 2025 at 5:48 PM
if you're in SF on monday, we'll have a hardware oriented meetup together with microsoft (right at the start of #msignite):
* Bad AI In A Box (BAIIAB) — making AI touch grass
* monitor our coffee consumption — I can now live debug coffee from across the world […]
[Original post on mastodon.social]
* Bad AI In A Box (BAIIAB) — making AI touch grass
* monitor our coffee consumption — I can now live debug coffee from across the world […]
[Original post on mastodon.social]
while we all love ourselves some ML, sometimes all you need is a little bit of math. great post from my colleague @alexandercanada on how to tune your shopping results based on history (with a function score)
https://alexmarquardt.com/personalizing-search-in-elasticsearch-without-ml-post-processing/
https://alexmarquardt.com/personalizing-search-in-elasticsearch-without-ml-post-processing/
November 12, 2025 at 7:34 AM
while we all love ourselves some ML, sometimes all you need is a little bit of math. great post from my colleague @alexandercanada on how to tune your shopping results based on history (with a function score)
https://alexmarquardt.com/personalizing-search-in-elasticsearch-without-ml-post-processing/
https://alexmarquardt.com/personalizing-search-in-elasticsearch-without-ml-post-processing/
got the right "business" cards for #kubecon
sorry, no YAML 😉
this was more fun to build than expected: @Claudeai generated HTML design with a *download as image* button. @moo to print and ship them
https://github.com/xeraa/code-card for the implementation
sorry, no YAML 😉
this was more fun to build than expected: @Claudeai generated HTML design with a *download as image* button. @moo to print and ship them
https://github.com/xeraa/code-card for the implementation
November 8, 2025 at 12:36 AM
got the right "business" cards for #kubecon
sorry, no YAML 😉
this was more fun to build than expected: @Claudeai generated HTML design with a *download as image* button. @moo to print and ship them
https://github.com/xeraa/code-card for the implementation
sorry, no YAML 😉
this was more fun to build than expected: @Claudeai generated HTML design with a *download as image* button. @moo to print and ship them
https://github.com/xeraa/code-card for the implementation
the #elasticsearch API docs alway s had cURL and console examples. a while ago we added dynamic languages (python, JS, ruby, PHP)
now, we have the first static language with java
big API surface so this will come in handy. even if just to tell an LLM "no, use […]
[Original post on mastodon.social]
now, we have the first static language with java
big API surface so this will come in handy. even if just to tell an LLM "no, use […]
[Original post on mastodon.social]
November 4, 2025 at 4:21 PM
the #elasticsearch API docs alway s had cURL and console examples. a while ago we added dynamic languages (python, JS, ruby, PHP)
now, we have the first static language with java
big API surface so this will come in handy. even if just to tell an LLM "no, use […]
[Original post on mastodon.social]
now, we have the first static language with java
big API surface so this will come in handy. even if just to tell an LLM "no, use […]
[Original post on mastodon.social]
#calhacks was a crazy weekend. great projects with @Elastic agent builder, RAG, geo and timeseries data: https://lnkd.in/dJ6H5cAk
PS: we also served a ton of (monitored) coffee. but that‘s for a separate post…
PS: we also served a ton of (monitored) coffee. but that‘s for a separate post…
October 27, 2025 at 5:40 AM
#calhacks was a crazy weekend. great projects with @Elastic agent builder, RAG, geo and timeseries data: https://lnkd.in/dJ6H5cAk
PS: we also served a ton of (monitored) coffee. but that‘s for a separate post…
PS: we also served a ton of (monitored) coffee. but that‘s for a separate post…
AWS outage / post-mortem: not one but two follow-up problems. and of course it started with DNS 😬
aws.amazon.com/message/101925/
aws.amazon.com/message/101925/
October 23, 2025 at 7:43 AM
AWS outage / post-mortem: not one but two follow-up problems. and of course it started with DNS 😬
aws.amazon.com/message/101925/
aws.amazon.com/message/101925/
#elasticon closing keynote by @swyx: the year agents ate search
from RAG, agentic RAG, and context engineering subagents 🔮
from RAG, agentic RAG, and context engineering subagents 🔮
October 21, 2025 at 11:24 PM
#elasticon closing keynote by @swyx: the year agents ate search
from RAG, agentic RAG, and context engineering subagents 🔮
from RAG, agentic RAG, and context engineering subagents 🔮
how much data can you get from coffee? a lot ☕️
this has been a fun project — from sending the data with OTLP to watching every step. can‘t wait to bring it to events and show streams of logs to dashboards
this has been a fun project — from sending the data with OTLP to watching every step. can‘t wait to bring it to events and show streams of logs to dashboards
October 20, 2025 at 4:30 AM
how much data can you get from coffee? a lot ☕️
this has been a fun project — from sending the data with OTLP to watching every step. can‘t wait to bring it to events and show streams of logs to dashboards
this has been a fun project — from sending the data with OTLP to watching every step. can‘t wait to bring it to events and show streams of logs to dashboards
tuesday Elastic{ON}: https://www.elastic.co/events/elasticon/san-francisco-ai
* agent builder, (vector) search performance, vibe coding,...
* @swyx will do the closing keynote for "the year agents ate search"
* I'm just putting the finishing touches on a […]
[Original post on mastodon.social]
* agent builder, (vector) search performance, vibe coding,...
* @swyx will do the closing keynote for "the year agents ate search"
* I'm just putting the finishing touches on a […]
[Original post on mastodon.social]
October 18, 2025 at 12:50 AM
tuesday Elastic{ON}: https://www.elastic.co/events/elasticon/san-francisco-ai
* agent builder, (vector) search performance, vibe coding,...
* @swyx will do the closing keynote for "the year agents ate search"
* I'm just putting the finishing touches on a […]
[Original post on mastodon.social]
* agent builder, (vector) search performance, vibe coding,...
* @swyx will do the closing keynote for "the year agents ate search"
* I'm just putting the finishing touches on a […]
[Original post on mastodon.social]
next week will be extra packed in SF
monday meetup: https://luma.com/smart-search
* ghita, ZeroEntropy: search tools for efficient AI agents
* jesse, fintools: LLMs and the next generation of financial search
* josh, reducto: building a vision-first RAG pipeline with reducto and elasticsearch
monday meetup: https://luma.com/smart-search
* ghita, ZeroEntropy: search tools for efficient AI agents
* jesse, fintools: LLMs and the next generation of financial search
* josh, reducto: building a vision-first RAG pipeline with reducto and elasticsearch
October 18, 2025 at 12:49 AM
next week will be extra packed in SF
monday meetup: https://luma.com/smart-search
* ghita, ZeroEntropy: search tools for efficient AI agents
* jesse, fintools: LLMs and the next generation of financial search
* josh, reducto: building a vision-first RAG pipeline with reducto and elasticsearch
monday meetup: https://luma.com/smart-search
* ghita, ZeroEntropy: search tools for efficient AI agents
* jesse, fintools: LLMs and the next generation of financial search
* josh, reducto: building a vision-first RAG pipeline with reducto and elasticsearch