UBI Assisted a Leading Pharmaceutical Company in Creating an AI-Based Search Engine to Search Within Organization Documents
Situation & Objective
The company sought a centralized search platform to efficiently search for content across multiple sources, including SharePoint, S3, and various web addresses. Their objective was to obtain accurate and relevant answers in a manner similar to Google search.
Solution Approach
To address this need, Useful Bi implemented a centralized data lake on AWS S3 using Python and Spark in Databricks. This data lake stored all the relevant documents along with their metadata.
Subsequently, AWS Kendra, an AI-powered search service, was leveraged to create an intelligent search engine.
Business Benefits Achieved
- Enhanced Search Efficiency : Employees experienced a 40% reduction in the time required to search for information.
- Improved Accuracy : 25% increase in accuracy of relevant answers to search queries.
- Centralized Search Platform : 30% improvement in productivity and collaboration. Users could access information from multiple sources through a unified interface.
