Designing AI-Assisted Procurement Platforms for Government Markets
American Journal of AI Cyber Computing Management · 30.01.2025 · 5 (4) · 58-64
DOI: 10.64751/ajaccm.v5i4.444 ↗Kurz gefasst
Fachartikel in der Zeitschrift American Journal of AI Cyber Computing Management, online erschienen am 30.01.2025. Verfasst von Hitesh Acharya. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0).
Abstract (Originalsprache)
Government procurement represents one of the largest commercial ecosystems globally, with the US federal government alone awarding over $700 billion in contracts annually. Despite this scale, contract discovery, requirement analysis, and bid preparation remain predominantly manual, creating barriers for small and mid-sized businesses and limiting lifecycle efficiency. This paper presents an architectural framework for AI-assisted procurement, drawing on the ContractSmart platform which facilitated over $2.3 billion in contract awards. We detail NLP pipelines for parsing government solicitations (SAM.gov, state portals), entity extraction for eligibility requirements (NAICS codes, set-asides, clearances), and recommendation algorithms matching contractor capabilities to opportunities. The platform reduced opportunity discovery time by 78%, improved bid-to-win ratios by 34%, and expanded identified pipelines by 4.2x. We address regulatory compliance (FAR/DFAR), small business preference preservation, algorithmic transparency, and bias monitoring for responsible AI deployment in government contexts
Zitieren
APA
Acharya, H. (2025). Designing AI-Assisted Procurement Platforms for Government Markets. American Journal of AI Cyber Computing Management, 5(4), 58-64. https://doi.org/10.64751/ajaccm.v5i4.444
Vancouver
Acharya H. Designing AI-Assisted Procurement Platforms for Government Markets. American Journal of AI Cyber Computing Management. 2025;5(4):58-64. doi:10.64751/ajaccm.v5i4.444
BibTeX
@article{Acharya2025,
author = {Acharya, Hitesh},
title = {Designing AI-Assisted Procurement Platforms for Government Markets},
journal = {American Journal of AI Cyber Computing Management},
year = {2025},
volume = {5},
number = {4},
pages = {58--64},
doi = {10.64751/ajaccm.v5i4.444}
}
EUVIMED