Blogsby GlyphNov 2
How our agent skills research, build, and test a complete Elastic integration
We pointed our agent skills at the NIST NVD API and got two data streams and a validated CEL program, with pipeline and system tests passing.
Elastic Search Labs the engineering blog of Elastic: latest posts and feed
Elastic · Engineering blog · 15 a month · 215 posts since Apr 20, 2022
Oct 30
We cut 40% of our MCP tools to get into Claude’s connector directory
Anthropic’s reviewers read every tool in our MCP connector and sent back fixes, wording changes and a case for publishing fewer of them, which is part of why mcp.elastic.co is now on the roadmap.
Oct 29
Elastic Cloud Serverless audit logs: Ship the trail to Observability or Security
Eight alerting rules and seven dashboards arrive with the logs, including Kibana Object Forensics, where a deleted detection rule shows up with the person who deleted it and the space it was in.
Oct 22
56 scanned pages to a cited answer in 5 minutes: RAG pipeline for the documents search cannot read
Ask a scanned, table-heavy PDF a question in plain language and get an answer stamped with the page it came from, because jina-ocr-v1 turns every page into structured Markdown that Elasticsearch embeds at ingest.
Oct 21
Bringing Elastic into Claude: The ‘why,’ the ‘how,’ and the ‘gotchas’
Point your Claude plugin at main and every merge ships as a release, and that’s the least surprising thing that happened between submitting and going live.
Oct 12
Your AI chat is not a screenshot: Live shared conversations in Elastic Agent Builder
Whoever picks up the problem next reads the whole thread and carries on with the same agent, with no access to anything beyond that one conversation.
Oct 8
How we’re building a context engine to cut agent token use by 71%
With a context engine, the same agent and model answered in 64 seconds using 9 tool calls. The baseline took 213 seconds, 33 calls, and missed a key relationship entirely.
Oct 5
The fastest work is the work you never do: 100x faster sorted queries in Elasticsearch
How Elasticsearch 9.6 lets ES|QL push its current TopN threshold into Lucene, and how that 100x shows up on some sorted queries and barely registers on others.
Oct 2
Check 100 candidates, not 10 million documents: Faster kNN filters in Elasticsearch
Elasticsearch now decides for each query whether to run a kNN filter before or after the vector search. On a 10M-vector corpus post-filtering was faster in 104 of 120 benchmark pairs while still returning k results.
Oct 1
Signed, sealed, delivered: ECK certificate management with Vault and cert-manager
Replace ECK's self-signed certificates with your enterprise PKI without renewing a certificate by hand, as Vault signs with a CA key sealed outside Kubernetes and cert-manager delivers fresh certificates to every Elasticsearch pod.
Sep 30
Using Jev as a search reranker: benchmarks and how to implement
We had Jev score Elasticsearch hybrid search results and let a short Python policy do the reranking, taking nDCG@10 from 0.9351 to 0.9565 on 250 Amazon Shopping Queries, and the code is all here.
Sep 29
How Elasticsearch Serverless hollow shards cut indexing-node shutdowns by 30%
Idle indexing shards in Elasticsearch Serverless now drop their Lucene writers and segment readers from memory until the next write, freeing indexing-tier heap at the cost of a median 218ms wait on that first write.
Sep 28
The best LLM writes correct Elasticsearch ES|QL 59% of the time. Here's what breaks the other 41%
We scored 6,000 ES|QL queries from four models against BIRD's answer key. Most misses come from mismatched join keys, SQL syntax the parser rejects, counting after a one-to-many join, or a value the model guessed.
Sep 25
GPU-accelerated vector indexing in Elasticsearch with NVIDIA cuVS: 138M vectors in under 10 minutes
Moving index builds to the GPU leaves the CPU free for queries, which is how vector indexing throughput went up 7x and p90 search latency fell 6x while indexing ran, with no change to recall.
Sep 24
Ask Elastic Agent Builder why it's slow: Natural-language trace analysis
Four agent performance questions your Agent Builder traces can answer, covering token spend by model, tool error rates, slow conversation turns, and recent prompts. The ES|QL for each is here, including the type cast SUM() needs.
Sep 21
Agentic workflows in Elasticsearch: pause an AI agent for human approval, resume 72 hours later
Build AI agent orchestration where the workflow waits for a human approval and then executes the fix on its own, with nothing extra to provision and the whole decision trail queryable in Elasticsearch.
Sep 18
AI video search with Elasticsearch and Jina: Find the exact seconds of footage you need
Cut each clip at its shot boundaries and embed every scene as a vector, and a plain text query gives back the file plus the exact seconds to drop on a timeline.
Sep 17
jina-ocr-v1: One OCR model for layout, tables, math and 100+ languages
jina-ocr-v1 scores 83.4 on olmOCR-bench with 570 million active parameters, the highest of any OCR model under 600 million, and it outscores GPT-5.2 on OmniDocBench.
Sep 16
One button, three places: How we rebuilt Kibana's page headers with stricter APIs
We gave Kibana's shared shell typed contracts, which is how design system governance became the default, and why the new page headers have no breadcrumbs.
Sep 15
Columnar storage isn't a columnar database. What Columnar mode brings to Elasticsearch
Elasticsearch has stored data in columns since 2013, but adding full columnar database capabilities required a new mode.
Sep 14
You and your AI agent shouldn't be using curl: Introducing the Elastic CLI and Agent Skills
Elastic CLI reaches every Elasticsearch, Kibana and Cloud API from one command, and it's what Elastic Agent Skills run on. Input is validated against a JSON Schema before anything leaves your machine, and API keys stay in your OS keychain.
Sep 11
Trust, but benchmark: How we let an AI agent optimize Elasticsearch
We share how we built a harness that automatically identifies and implements optimizations in the Elasticsearch codebase.
Sep 11
How we built PromQL into Elasticsearch
PromQL runs on the same Elasticsearch compute engine as ES|QL, with no plugin and no separate process to operate. Getting there meant changing how the engine evaluates time windows and builds grouping keys.
Sep 9
Elasticsearch Vector Database: Ship in minutes, scale affordably to hundreds of billions
The hard parts of hybrid retrieval, already done, with optimized defaults, third party and native Jina AI models, and managed GPU inference all out of the box. Build fast, scalable AI apps, not infrastructure.
Sep 8
One field, one copy: How Elasticsearch columnar storage drops the inverted index
Storing each field once means no inverted index, so doc values now read in bulk and skippers let queries skip whole ranges of documents, while new mapping attributes control what each field is allowed to contain.
The latest 200 of 215 posts on file; older ones are under Months.
About this blog
Vector search, RAG, relevance and Elasticsearch internals.
Posts per week, last 26 weeks
15 a month over the last six months (or since the feed's oldest post)
Posts by month
Writes about
AI and agents (26), Databases (21), Reliability and performance (16) in the last six months.
Elastic
Elastic on Companies (companies.fru.dev)Across Glyph:Acquisitions·advisories·awards·breaches·calendar·Attested·Quarterly·Downtime·Leaders·marketplaces·Allied·Paydays·Metered·problems·quadrant·Changefeed·repos·salaries·Nines·socials·standards·Stories·sunsets·TechConf·terms·Titlewave
History
Official checkSep 25
not reviewed to official: on elastic.co