Elastic

What They Build

Elastic makes Elasticsearch — the dominant open-source distributed search engine built on Apache Lucene. It is the default backend for a large share of production search systems globally (e-commerce, log analytics, enterprise search). Elastic also runs a cloud offering (Elastic Cloud) and a Search Labs blog that publishes practical search engineering content.

Search Contributions

Retrieval Models

• ELSER (Elastic Learned Sparse Encoder) — production-grade sparse retrieval model based on SPLADE. Achieves +17% NDCG@10 over BM25 on BEIR benchmarks with zero-shot performance

• BBQ (Better Binary Quantization) — scalar + binary quantization approach for Elasticsearch dense vector fields; competitive with TurboQuant at fraction of the cost

• Block-Max WAND — introduced in Lucene 8.0 / Elasticsearch 7.0 (2019) by Adrien Grand; 3x–15x speedup for top-k retrieval; required API change (hits.total now object with value+relation, controlled by track_total_hits)

Search Features

• function_score query — the primary Results Boosting mechanism in Elasticsearch
• kNN + pre-filter pipeline — Vector Filtering support in dense vector search
• Native Hybrid Search with BM25 + dense vector combination

Applied Examples Published

• Multilingual hybrid search + reranking on Elasticsearch
• E-commerce personalization via purchase history in Elasticsearch
• Margin and popularity boosting patterns (governed control plane)

Tech Stack Significance

Elasticsearch is used as the underlying engine in many case studies in this vault:

• Zalando — Base Search layer
• Uber — custom Lucene platform
• Semantic Scholar — Elasticsearch with ML reranker on top

People

• Thomas Veasey — Principal Engineer, Elastic

• Leonie Monigatti — Developer Advocate / ML Engineer; writes on agentic search and context engineering

• Adrien Grand — Principal Engineer, Elastic; core Lucene committer; implemented Block-Max WAND

Articles

• Elastic Learned Sparse Encoder ELSER Retrieval Performance

• Elasticsearch BBQ Optimized Scalar Quantization vs TurboQuant

• Multilingual Embedding Model Hybrid Search Reranking

• Ecommerce search optimization with margin & popularity boosting

• Late Interaction Models - How to Scale and Optimize in Elasticsearch

• Agentic Search for Context Engineering — Leonie Monigatti; three search tool patterns for context engineering

• Faster Retrieval of Top Hits in Elasticsearch with Block-Max WAND

Concepts

ELSER · BBQ · BM25 · Block-Max WAND · WAND · Results Boosting · Hybrid Search

Governed Ecommerce Search Series (2026)

A 7-part series on building governed search control planes for ecommerce:

• Why Ecommerce Search Needs Governance and How It Improves Retrieval — intent classification + routing
• Ecommerce Search Governance - Move Faster Not Slower — zero-deploy operating model
• Elasticsearch Personalized Search in Ecommerce - Improve Relevance — purchase history + cohort personalization

Additional People

• Alexander Marquardt — Elastic Services Engineering
• Honza Král — Elastic
• Taylor Roy — Elastic
• Peter Straßer — Elastic; late interaction / ColPali scaling
• Benjamin Trent — Elastic; late interaction / ColPali scaling

Additional Concepts

Search Governance · Personalization

Tools

• Elasticsearch