[For more diverse corporate information on the startup mentioned in this article, please visit Unicorn Factory's big data platform 'Data Lab'.]
On the 5th, AI platform specialist Infinic announced that it has begun developing an intelligent data processing system to convert handwritten military logistics documents into AI-based digital data.
Infinic completed its initiation report on 'Development of an OCR-based intelligent data processing and structuring system for unstructured handwritten documents in the military logistics sector' as part of the Ministry of SMEs and Startups' 2026 second round SME (small and medium-sized enterprise) technology development support project, with a target delivery date set for December. The core objective is to automatically recognize unstructured documents at military logistics sites—such as supply requisition forms, maintenance records, and material status tables—using AI-powered optical character recognition (OCR), then structure them into data compatible with the Defense Logistics Information System (DELIIS).
Currently, most forms used in military logistics operations are filled out by hand and subsequently re-entered manually by staff into computer systems. General-purpose OCR based on external cloud APIs faces limitations in recognizing military-specific codes such as National Stock Numbers (NSN) and unit abbreviations, making them difficult to apply in actual field operations.
To address this challenge, Infinic plans to develop a three-axis recognition engine that integrates high-precision character recognition (OCR), context understanding based on vision-language models (VLM), and ontology-based logical validation. In this system, OCR reads handwritten characters and special codes; VLM corrects misrecognitions based on contextual analysis; and the ontology verifies data consistency according to military logistics domain rules. The company aims to achieve a character recognition rate of over 95%, a character error rate (CER) below 5%, and processing time under five minutes per document.
The system is designed to operate independently within a closed network completely isolated from the external internet, featuring a five-layer security framework that includes data encryption, role-based access control, and automatic personal information masking. The company expects that once deployed, personnel currently engaged in manual data entry can be reassigned to analytical and decision-making tasks, while reducing data input and verification time by more than 50%.
To ensure technical maturity and domain suitability, the Korea Military Academy will participate as an advisory institutional investor. The academy will provide development space within its research facilities and jointly conduct ontology design and validation throughout the project period.
Infinic plans to link the structured military logistics data with its own ontology-based decision support platform, 'AURON'. This integration will enable semantic relationship-based search and retrieval of maintenance histories and material status information, as well as analysis of impact scope and maintenance priority based on relationships among equipment, units, and materials. In the future, the company intends to visualize this data via dashboards and expand the system into a military-specific AI platform usable across units, facilities, and job functions.
An Infinic representative stated, "This project goes beyond digitizing handwritten documents; it is part of the defense digital transformation process that converts military logistics information into data assets." He added, "Based on data processing technologies tailored to the military environment, we plan to continue expanding our digital transformation initiatives in the defense AI and public sectors."
[MoneyToday startup media platform Unicorn Factory]
in FUTURE
