←
Elastic FR4,036 posts
Elastic FR avatar

Elastic FR

@ElasticFR

Suivez-nous pour être au courant de ce qui se passe au niveau de la communauté francophone elastic. Suivez @elastic pour des informations sur #ElasticStack.

◷ Joined November 2011▧ 806 Photos and videos
Elastic avatar
⟳ Elastic FR retweeted
Elastic@elastic·Jun 22

3 patterns for multimodal RAG. Here's how they differ and when each one breaks down. Most RAG systems add multimodal support by converting everything to text first. Is your system natively multimodal, or just a conversion pipeline? The architecture choice shapes what you can query and what you lose. Shared vector space - Cross-modal search without format conversion - Requires large multimodal training datasets - Semantic drift is a real risk if training data is narrow Single grounded modality - Works with any existing text search setup - Spatial relationships in images don't survive conversion - Retrieval quality depends on captioning/transcription accuracy Separate retrieval pipelines - Best per-modality retrieval accuracy - Most complex to rank across modalities - Highest compute cost, independent search per modality Pick your pattern, clone the repo, and build it.

Post attachment
Elastic avatar
⟳ Elastic FR retweeted
Elastic@elastic·May 19

Exact kNN on 1M vectors: ~10 QPS HNSW on the same dataset: ~1000 QPS This is because HNSW builds a layered graph: shortcuts at the top, all vectors at the bottom. A query enters at the top layer, finds the closest node, uses it as the entry point for the layer below, and repeats until it reaches a tight local search at layer 0. It also let you tune your indexing speed vs query speed vs recall, so you can get the index that suits your needs.

0:26
Elastic avatar
⟳ Elastic FR retweeted
Elastic@elastic·May 8

30x faster than Prometheus on gauge averages sounds like a typo. It's not. Elasticsearch 9.4 rebuilt TSDS and ES|QL into a fully columnar metrics engine. Vectorized time series aggregations, zero-copy decoding, doc value skippers replacing inverted indices. The result at 1.4M time series: queries that took Prometheus 30+ seconds finish in under 2.

Post attachment
Elastic Dev avatar
⟳ Elastic FR retweeted
Elastic Dev@elastic_devs·Apr 10

semantic_text now defaults to jina-embeddings-v5-text on Elastic Inference Service. Multilingual semantic search. Zero extra configuration. Define the field, index your data, search. Query in English: "comfortable headphones for long flights" A French review ranked above the English one, because "très confortable pour les longs trajets en avion" was a closer semantic match than the English review, which split its focus across battery life and noise cancellation.

Post attachment
Elastic avatar
⟳ Elastic FR retweeted
Elastic@elastic·Mar 20

Elastic is heading to KubeCon Amsterdam 2026 🇳🇱 Visit booth #980 to see how we’re building OpenTelemetry to simplify ingestion, reduce pipeline complexity, and deliver consistent observability across metrics, logs, and traces with a focus on agentic observability. Come see the demos and meet the team: go.es.io/4uGtF3N

0:43
Elastic Dev avatar
⟳ Elastic FR retweeted
Elastic Dev@elastic_devs·Mar 4

Don’t .query() me twice! Yours truly

1:53
Elastic Dev avatar
⟳ Elastic FR retweeted
Elastic Dev@elastic_devs·Mar 4

APT confirmation used to take hours. Now it takes 4 minutes. Attack Discovery correlates alerts into a single narrative. A workflow triggers the agent. The agent: • Looks up the hash on VirusTotal • Runs ES|QL queries across your logs • Finds the on-call analyst • Creates a case • Opens a Slack incident channel All before you read the threat intel report.

Post attachment
Elastic Dev avatar
⟳ Elastic FR retweeted
Elastic Dev@elastic_devs·Feb 25

Nothing to buy, no license to upgrade, and no approval process.

Attention Elasticsearch Community! Elastic AutoOps is now free for every self-managed cluster with no license upgrade needed. AutoOps analyzes your clusters, identifies what is wrong, and tells you how to fix it: ✅ Automatic issue detection + fix recommendations ✅ Real-time slow query and bottleneck detection ✅ Resource optimization insights ✅ Central view of all clusters, no extra monitoring setup Learn more → go.es.io/40yKlME

0:19
Elastic Dev avatar
⟳ Elastic FR retweeted
Elastic Dev@elastic_devs·Feb 24

Forcing LLMs to follow a strict schema is hard. Build reliable AI agents using structured outputs This guide shows how to ground your agents using data contracts to ensure valid responses every time. What you will learn: • How to define your schema for agentic workflows • Using the Elasticsearch inference API for data extraction • Implementing validation Stop guessing if your agent will return valid JSON.

David Pilato 🇺🇦🇪🇺🇫🇷 avatar
⟳ Elastic FR retweeted
David Pilato 🇺🇦🇪🇺🇫🇷@dadoonet·Jan 26

Et voilà ! So proud to reach the 100th meetup milestone for the @ElasticFR 🇫🇷 community. A huge thank you to our speakers, sponsors, and community members for making every event a success! ❤️ 📺 Watch here: youtube.com/watch?v=GYDkfGDJ…

Elastic Dev avatar
⟳ Elastic FR retweeted
Elastic Dev@elastic_devs·25 Dec 2025

What are you going to eat today? For the final advent calendar gift of the year, @PiotrPrz shares a simple 3-step script to get the healthiest possible food, whatever language you are using, thanks to our OOTB JINA model: go.es.io/4qiscxB

Post attachmentPost attachmentPost attachment
Elastic Dev avatar
⟳ Elastic FR retweeted
Elastic Dev@elastic_devs·15 Dec 2025

We have added a new and interesting retriever! The `diversify` retriever reduces the result set from another retriever by applying a diversification strategy to the top-N results, helping prevent similar documents from dominating the top results.

Elastic Dev avatar
⟳ Elastic FR retweeted
Elastic Dev@elastic_devs·14 Dec 2025

The advent calendar post of the day is all related to the powerful maps feature in #Kibana. Combine it with ES|QL and 🥳

Post attachment