Oracle improves employee intranet experience with OCI Generative AI

Oracle powers its centralized intranet with OCI Generative AI for enhanced clickthrough rates and a better user experience.


The release of MyOracle Search with semantic search and OCI Generative AI significantly enhanced the search experience for all employees. Clickthrough rates improved from about 40 percent to about 70 percent and we saw improved content comprehension.

Sudhir DurejaSenior Director of Engineering, Oracle Cloud Infrastructure (OCI)

MyOracle, Oracle’s centralized intranet, receives over five million monthly hits from remote and office-based employees. One of the portal’s key features is MyOracle Search, an enterprise search service that gets over 500,000 queries per month. Previously, MyOracleSearch had a 30% to 40% clickthrough rate due to limitations with traditional lexical search, such as lack of semantic understanding, relevance ranking, natural language processing, and ability to capture intent. To improve the user experience, the search team implemented Search Generative Experience (SGE) on Retrieval Augmented Generation (RAG) architecture using OCI Generative AI, OCI Data Science platform, and OCI Open Search. Instead of simply displaying a list of records found, as traditional searches would, the RAG architecture uses a text generation model to produce a natural language response that answers the original question, with those responses grounded in documents from the Oracle knowledgebase to improve accuracy. To help users further, the generated answer is supported by references along with links to the source documents. This feature helped lift clickthrough rates to about 70% and garnered over 75% positive employee feedback on the quality of generated answers.

Published:April 16, 2024