Kyungbang

Kyungbang, a global textile manufacturer, manages one of the most complex challenges in its industry: high-volume fiber testing across suppliers and factories worldwide. Data arrived in scattered formats—PDFs, Excel sheets, raw text—forcing engineers to spend hours reconciling numbers before making decisions. Edge8 partnered with Kyungbang to build an AI-powered quality platform designed to unify that data, flag anomalies in real time, and allow engineers to query results in natural language.

• Supplier and factory data scattered across inconsistent formats (PDF, Excel, text)
• Manual entry of results into spreadsheets, then printed for review
• No direct connectivity from testing machines to decision-makers
• Engineers lacked intuitive tools to quickly query or compare test outcomes

• Built an AI-driven document parsing engine to ingest QC reports from any format
• Developed a system for real-time anomaly detection and automated flagging
• Enabled natural language querying for batch, machine, and location histories
• Integrated the system directly into existing QC workflows across facilities

• Reduced manual reconciliation time by over 50%
• Engineers retrieve insights instantly using everyday language instead of spreadsheets
• Faster, more confident production decisions across multiple facilities
• Established a scalable foundation for predictive blending and future AI-driven enhancements