Enhancing Content Findability with LLM-powered Solutions

This case study details how a USA-based SaaS company implemented SearchUnify’s LLM-powered Generative Question Answering feature to address challenges related to content findability and agent productivity. The company faced issues with fragmented articles and a lack of personalization, which resulted in low case deflection rates and increased resolution times. By mapping Generative Question-Answers to specific articles and utilizing a dataset similar to SQUAD for training Language Models, SearchUnify improved the relevance of knowledge base information. The integration of this feature led to significant improvements in customer effort scores, reduced support costs, and increased self-service resolution rates. Specifically, support costs decreased from over $560,000 to $310,000 within three months, and the self-service rate rose from 78% to 89%. The study highlights the effectiveness of personalized content delivery in enhancing user experience and operational efficiency.