Broadcasting Media Equipment Industry Center Unveils Public Procurement Monitoring System

The Korea Radio Promotion Association's Broadcasting Media Equipment Industry Center has developed and launched a monitoring system that automatically tracks the entire public procurement process for video and audio equipment using generative AI-based Vibe Coding technology. Named the 'Broadcasting Media Equipment Management System', it integrates with the Public Procurement Service's KONEPS (Korea ON-line Public E-Procurement System) API to automatically collect information on bidding plans, pre-specification disclosures, tender notices, and contract awards for 98 equipment items, and includes a function to identify notices with missing item codes. The system offers features for opinion management on pre-specification notices, tracking acceptance status by procuring agencies, management of review committees and evaluations, and weekly/monthly result reports (exportable to Excel/PDF). General users can utilize four monitoring menus. The operation of this system is expected to improve the time and effort previously spent manually searching for related projects on KONEPS, managing approximately 1,000 cases of broadcast equipment procurement information annually.

Seo Chang-ho, a RAPA executive who developed the system, stated that it was built without a separate development budget by utilizing AI coding tools like Claude Code.

A key challenge in operating the system is cost. Advanced features such as automatic result report generation utilize APIs from commercial AI model services like OpenAI or Claude, and costs increase with usage. As the system is operated for public interest and excludes profit models, the external disclosure of result report functions is on hold due to the cost burden of using commercial AI API services.

Seo highlighted the free implementation of the practical system using AI coding tools such as Claude Code and proposed policy support based on proprietary AI models, such as the 'Dokpamo' project results from the Ministry of Science and ICT, due to concerns over increasing AI API costs. He suggested that once proprietary AI models, which are the results of the Ministry of Science and ICT's 'Proprietary AI Foundation Model Project (Dokpamo)', are commercialized, policies should be prepared to support AI model API usage for public interest activities.