Elastic (@elastic) / X

Elastic

10.4K posts

Elastic

@elastic

Where developers learn, build, and share. Your source for hands-on demos, cheat sheets, explainers and more.

Global

Joined October 2009

Elastic
@elastic
12h
Hybrid search = BM25 + vector search, merged and reranked in 1 request. Most search implementations pick one. Lexical search misses semantic intent. Vector search misses exact keyword matches. Hybrid covers both. Here's how it fits together in a single Elasticsearch query: -
1.5K
Elastic
@elastic
12h
Resource:
What is hybrid search? How it works and when to use it
From elastic.co
387
Elastic
@elastic
Jun 25
Your Claude Code agent forgets everything between sessions. So you bolt on a memory service. Another API, another thing to run. If you already run Elasticsearch, you already have the parts. semantic_text handles embeddings at index time. ES|QL gives you hybrid recall in one
2.8K
Elastic
@elastic
Jun 25
Full walkthrough, schema, and the bridge CLI:
Persistent memory for agents: Claude Code on Elasticsearch - Elasticsearch Labs
From elastic.co
600
Elastic reposted
JP Hwang
@_jphwang
Jun 25
Cooking up a new video + a live session to demo video search with @JinaAI_ omni v5 model and @elastic. Here's a preview of the demo app running entirely locally & live on my Mac. More soon 😉
00:00
820
Elastic
@elastic
Jun 24
Most data analysis still feels like translation work. Someone asks which products drive revenue. Then you write queries, join tables, check dashboards, validate assumptions. An hour later, maybe you have an answer. This tutorial wires Elastic Agent Builder MCP to @awscloud
GIF
1.7K
Elastic
@elastic
Jun 24
Tutorial and full code: go.es.io/4uVdj6j GitHub repo: go.es.io/4wdcEyl
Elastic Agent Builder MCP and AWS AgentCore: Python tutorial - Elasticsearch Labs
From elastic.co
677
Elastic
@elastic
Jun 23
Most search benchmarks only tell half the story. You test relevance. You ship it. Then p99 latency tanks under real concurrency and users start filing tickets. Or you optimize for speed, and your top-k results are fast garbage. The fix: measure both sides every time. 10
1M
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
2.5M
Elastic
@elastic
Jun 22
Learn more:
Multimodal RAG: Building a multimodal RAG system with Elastic - Elasticsearch Labs
From elastic.co
967
Elastic
@elastic
Jun 19
Run the same vector search benchmark across 3 engines. Jingra is a new open-source framework that uses YAML config to run identical workloads on Elasticsearch, OpenSearch, and Qdrant. The part that matters: with await_index_ready enabled, it holds evaluation until the index
2.2K
Elastic
@elastic
Jun 19
Blog: go.es.io/4vXMUpl Repo: go.es.io/4oNsPjd
Vector search benchmarking with Jingra: A reproducible framework - Elasticsearch Labs
From elastic.co
1K
Elastic
@elastic
Jun 18
0.89 recall at k=10 and zero cross-tenant leaks on a persistent agent memory layer. Built on Elasticsearch with 3 indices mapped to cognitive science: episodic events, semantic facts, procedural playbooks. Each has its own write rate, aging rules, and update logic. Episodic
4.6M
Elastic
@elastic
Jun 18
Blog and full implementation: go.es.io/4oFWhYy GitHub repo: go.es.io/4uKA8tl
Agent memory on Elasticsearch: hybrid retrieval and DLS - Elasticsearch Labs
From elastic.co
1.1K
Elastic reposted
Elastic Security Labs
@elasticseclabs
Jun 17
Live now!
REPLAY
Elastic Security Labs
@elasticseclabs
Put agentic AI to work: Real-world defense against threats
1.1K
Elastic
@elastic
Jun 17
🧵 We've reworked stored-vector kNN search to cut latency by up to 3x with a single API change. Here's why 2 round trips was always unnecessary, how the speedup works, and what the benchmark numbers show.
1.6K
Elastic
@elastic
Jun 17
Replying to @elastic
3/ In 9.4, this collapses into a single request: Elasticsearch fetches the stored vector internally and uses it directly. One request. No client-side plumbing.
517
Elastic
@elastic
Jun 17
4/ We ran this on 2M documents across 2 nodes in the same GCP zone. Same data center. Same availability zone. Still: p50: 10.4ms → 3.1ms (3.3x faster) p90: 25.4ms → 5.9ms (4.3x faster) p99: 27.7ms → 8.1ms (3.4x faster) Even when nodes are physically close,
Up to 3x faster stored-vector queries in Elasticsearch - Elasticsearch Labs
From elastic.co
437