SAT Information Co.,Ltd.

Oh Sae Yong, CEO

In the era of generative AI in 2025, the safe utilization of text-based unstructured data has emerged as a key challenge. This presentation will introduce secure preprocessing and utilization strategies for AI training data, centered on SAT Information’s 15 years of accumulated expertise in electronic document analysis and intelligent text parsing technologies.

In particular, it will present a practical framework that combines unstructured-to-structured data transformation with automated personal data detection and de-identification technologies. This enables organizations to produce high-quality data while complying with personal data protection regulations in RAG (Retrieval-Augmented Generation) environments.