Flyle Inc., a provider of AI transformation solutions for contact centers and CX, announces the case study and results of its utilization by Orient Corporation (hereinafter referred to as Orico). Previously, Orico had approximately three staff members manually extracting and classifying customer opinions and requests from contact history entered by operators. This internal utilization was limited, and there was a challenge in fully connecting the classification results to improvement measures. In 2025, Orico introduced Flyle, which automatically classifies and visualizes monthly logs of 150,000 conversations using generative AI. This has reduced the classification work, equivalent to approximately 7,350 hours per month, to nearly zero. It is also being used to identify precursors to complaints and customer harassment, and to extract insights for improvement. ■Background of Flyle Introduction Orico has established "customer-centricity" as its basic management policy and is advancing the creation of a system to implement a customer perspective across the entire company, including the reorganization of the former CX Promotion Department into the CX Management Department in April 2026. On the other hand, there was no established mechanism to quantitatively organize customer feedback and share it as factual data, making the creation of a foundational system to grasp the overall picture a major goal. The gap with companies that have been actively utilizing customer feedback for a long time was widening, leading to a growing sense of crisis that they would "no longer be chosen" by customers if this continued. The turning point came in January 2025 when the transcription of conversation logs began, leading to the accumulation of data on the scale of 150,000 conversations per month. However, the text data from conversation logs did not always have high accuracy, and conventional text mining tools, which excel at keyword extraction, could not understand the context and visual