Showcase Begins Proof-of-Concept for Next-Generation Masking Technology Using LLM in its "ProTech AI-OCR" Image Analysis Service
Showcase Inc. has initiated a proof-of-concept for high-precision masking technology using Large Language Models (LLM) within its "ProTech AI-OCR" image analysis service. This new technology aims to accurately mask information in unstructured documents and contexts requiring comprehension, which were challenging for conventional OCR, thereby improving operational efficiency and advancing information management.
📋 Article Processing Timeline
- 📰 Published: May 18, 2026 at 20:00
- 🔍 Collected: May 18, 2026 at 11:31
- 🤖 AI Analyzed: May 18, 2026 at 19:32 (8h 1m after Collected)
Showcase Inc. (Minato-ku, Tokyo; President: Koichi Matsumoto; Securities Code: 3909; hereinafter "Showcase"), which develops DX cloud services connecting companies and customers, announces that it has begun a proof-of-concept for high-precision masking technology using Large Language Models (LLM) in its image analysis service "ProTech AI-OCR." Our image analysis service "ProTech AI-OCR" is a function that uses AI machine learning to automatically mask personal and confidential information contained in various documents. Previously, we have improved accuracy by individually training new patterns for cases where masking precision was insufficient. However, a challenge remained with document formats that differ for each municipality or company, and cases requiring information judgment based on context, which were difficult for conventional learning-based OCR to handle. To solve this issue, we have been promoting research and development of a new technology that utilizes Large Language Models (LLM) to identify and mask information based on contextual understanding. As a result, we have achieved a technology that can mask with high precision even on unstructured documents that were difficult for conventional methods, and can automatically mask a wide range of information regardless of document format or image quality. Currently, this technology is being pilot-tested in some municipalities for the purpose of masking "credit card numbers," and we are proceeding with verification to improve operational efficiency and advance information management. We will continue to aim for further accuracy improvements and expansion of application areas, and strive for utilization in a wide range of fields by providing solutions that support safe and efficient information management. Not just character analysis, the next-generation OCR "ProTech AI-OCR" "ProTech AI-OCR" is a service that can read and textify character information from all kinds of images, such as ID cards like driver's licenses, business cards, receipts, as well as handwritten characters on application forms and medical questionnaires, with an accuracy of over 99%. This automates the previously manual input work, reducing work time and costs. Service site here: https://ekyc.showcase-tv.com/ai-ocr/ Contact us about the service here: https://ekyc.showcase-tv.com/ai-ocr/contact/ 【Showcase Inc. Company Profile】 President: Koichi Matsumoto Location: Roppongi First Building 14F, 1-9-9 Roppongi, Minato-ku, Tokyo Established: February 1, 1996 Stock Exchange Listing: Tokyo Stock Exchange Standard Market Securities Code: 3909 Corporate Site: https://www.showcase-tv.com/ Core Value: "Making people happy with Omotenashi Technology" We help remove the "inconveniences" people feel and guide them to "happiness." And our joy is to increase the number of happy people with Omotenashi Technology, a mechanism that exceeds people's expectations and provides excitement and satisfaction.