Artificial intelligence (AI) agents will be introduced to public software (SW) project ordering tasks. As AI assists from drafting Requests for Proposals (RFPs) to reviewing compliance with laws and regulations, preparing cost breakdown statements, and supporting task deliberation, the AI Transformation (AX) in the public procurement sector is expected to accelerate.
According to the industry on May 25, the Public Procurement Service (PPS) and the National Information Society Agency (NIA) have embarked on building the “Public SW AI Accountable Project Management Platform.”
This project is an extension of the “Measures to Lead the Activation of the AI Industry in Public Procurement” announced at the Economic Ministers' Meeting last February. The PPS plans to first introduce and pilot-operate AI agents in the public SW ordering domain.
A project insider said, “Since this is the first project, testing and verification are necessary, but if performance is confirmed, it can be expanded to other procurement projects.”
The core of this platform is the establishment of an “AI-based Procurement Support Service.” Utilizing generative AI, the platform provides procurement officials with customized draft RFPs by project type, analyzes requirements included in the RFP, and drafts cost breakdown statements based on Function Points (FP) and Man-Months (M/M). It also self-diagnoses potential violations of relevant laws and systems, such as the Software Legislation Promotion Act, the Electronic Government Act, and the Act on Contracts to Which the State Is a Party, and guides on matters that require modification during the procurement request process.
The tasks of the Task Deliberation Committee will also be digitized. Based on relevant data such as task deliberation requests, RFPs, and cost breakdown statements, it supports the drafting of deliberation forms and notification of results, and assists the public SW project management procedures, such as calculating appropriate project periods, based on AI. In addition, the PPS plans to prepare a small Large Language Model (sLLM) specialized for procurement tasks.
Public SW procurement is a complex task that requires the simultaneous review of numerous laws and guidelines. If AI agents take charge of drafting procurement proposals, checking regulations, preparing cost breakdown statements, and supporting task deliberation, it is projected to reduce variance in officer competency and mitigate project delays. It is expected to shorten task processing times, such as generating and reviewing RFPs, by 30 to 40% and reduce administrative costs by 20 to 25%.
As the government is paying close attention to the AX transformation of the public procurement market, if the public SW procurement management platform successfully settles into the actual field, the AX transformation across the entire public procurement market is expected to gain momentum. Previously, the PPS established the “Master Plan for Public Procurement AI Transformation” and materialized plans to introduce AI services at each stage of public procurement. The service also plans to build an “Intelligent Public Procurement Platform” that integrates AI procurement support tools next year.
