Enhancing Content Findability with LLM-powered Solutions

This document is a case study detailing how a USA-based SaaS company improved content findability and agent productivity through the implementation of SearchUnify’s LLM-powered Generative Question Answering feature. The company faced challenges with its online customer community, where users often encountered irrelevant search results and support agents struggled to find pertinent information, leading to low case deflection rates. The SearchUnify team addressed these issues by mapping Generative Question-Answers to specific articles and training Language Models (LLMs) on a relevant dataset. This approach enhanced the relevance of knowledge base information and personalized the user experience based on their profiles. The integration of this feature resulted in significant improvements, including a reduction in support costs from over $560,000 to $310,000, an increase in customer effort score from 62% to 84%, and a self-service resolution rate that rose from 78% to 89%. Overall, the case study outlines the successful application of LLM technology to enhance user experience and operational efficiency.