Case Study: LG CNS achieves 95% search relevance and halves data retrieval time with Elastic (Elasticsearch)

LG CNS implements Elasticsearch hybrid, vector, and generative AI capabilities to improve accuracy and reduce data retrieval times

LG CNS, the IT services arm of LG Group, set out to improve its in-house KeyLook AI retrieval-augmented-generation (RAG) search system after keyword-based search struggled to capture user intent, handle synonyms, typos and cross-language queries. The challenge was to move from traditional full-text search to a context-aware vector approach that could scale for large corporate knowledge-management (KM) datasets and deliver faster, more accurate answers.

By integrating Elasticsearch’s hybrid search capabilities (full-text, sparse and dense vector, and semantic search), LG CNS boosted search relevance from 75% to 95% and cut mass data retrieval time in half (from 0.2s to 0.1s) while encoding and indexing corporate documents for generative-AI answers. The solution also improved KM usability and security controls, is being piloted internally, and lays the groundwork for multilingual support and expanded next‑generation KM services.

LG CNS

Youngmin Kim

AI Lab Language General Consultant