LG CNS improves accuracy and reduces data retrieval times by 50% with Elastic

Improving search accuracy and performance for knowledge management

LG CNS is Korea's leading IT services company. It provides consulting, system integration, network integration, and business-process outsourcing services. LG CNS faced challenges in enabling efficient, accurate, and scalable search functionality for its KeyLook AI algorithm.

Traditional keyword-based search methods were limited in understanding user intent, leading to irrelevant results. This hindered knowledge management (KM) initiatives and required significant manual effort to refine searches. LG CNS needed to process massive datasets quickly, which strained its existing infrastructure and led to inefficiencies in memory expansion and data retrieval speeds.

Reducing mass data search time by more than half

By adopting the Elasticsearch Platform, LG CNS revolutionized its search capabilities. Integrating Elastic’s sparse and dense vector search modules with LG CNS’s AI models enabled LG CNS to increase search accuracy from 75% to 95%. Hybrid search allowed the models to understand user intent better and return relevant results, even for queries with typos or synonyms.

LG CNS reduced mass data retrieval times by over 50% by leveraging our scalable and optimized architecture, enabling faster access to critical information. Enhanced knowledge management tools empowered employees with a seamless and secure search experience, significantly improving efficiency and usability.

"The integration of the KeyLook algorithm with the Elasticsearch Platform has been instrumental in enabling us to search and utilize the knowledge we need in a GenAI-powered conversational format."

— Youngmin Kim, LG CNS AI Lab Language General Consultant

Supercharging knowledge management with the Elasticsearch Platform

The Elasticsearch Platform was the ideal solution for LG CNS's requirements, offering advanced vector search capabilities, robust memory management, and high-speed data retrieval. By encoding corporate data using generative AI models and indexing it within the platform, LG CNS ensured accurate, context-aware search results. Hybrid search functionality enabled combining full-text, vector, and semantic search methods to optimize relevance and intent recognition.

On-premises deployment of the Elasticsearch Platform further improved the performance of the KeyLook AI algorithm, enabling real-time retrieval and integration with ChatGPT for conversational AI-powered responses. The system enhanced LG CNS's KM model, delivering concise, actionable insights while maintaining strong security protocols for access control and filtering.

We provided LG CNS with an adaptable, scalable, and reliable solution for its AI needs. Our unique capabilities, including hybrid search modules, real-time aggregation, and seamless integration with generative AI models, allowed LG CNS to significantly enhance search relevance and performance.

"We found that only Elastic supported our semantic-search use cases. The sparse vector search developed during our research showed very good performance, and we wanted a search engine that could be equipped with it," says Kim. "Elastic's sparse vector module allows us to perform searches even when keywords are not exact matches and when they include words with similar meanings or contain typos. The results of adopting Elastic with the help of our partners have been remarkable."

The Elasticsearch Platform also supports LG CNS's innovation in knowledge management by enabling efficient indexing, memory optimization, and multi-language support for expanding global markets. With the Elasticsearch Platform, LG CNS was able to transform scattered information into valuable assets, positioning it as a leader in next-generation knowledge management services.