AI predictive model for public procurement
Industria
Government
Escenario
BI and AI
Cliente
Government Entity
As part of an initiative to enhance transparency and efficiency in public procurement processes, the Public Procurement Observatory was developed. This project was funded by the International City/County Management Association (ICMA) after winning a U.S. Government competition. Arkkosoft was responsible for implementing the solution for a government entity, structuring the project into two phases.
The first phase involved creating a BI-based portal, leveraging Azure Power BI, and a WordPress-based platform for data publication. In the second phase, AI models were developed using Azure Machine Learning to analyze procurement processes, identifying whether contract amounts exceeded previous values or if a tender might remain unawarded. Additionally, an AI model was incorporated to calculate the probability of procurement objections, optimizing decision-making and strengthening transparency in public administration.
Solution Components:
Azure AI (Artificial Intelligence):
Implementation of traditional Machine Learning (ML) models to predict contract amounts and procurement objections.
Azure Data Factory
ETL (Extract, Transform, Load) process management to integrate data from multiple sources.
Intermediate Databases on Azure:
Efficient storage and processing of information.
VPNs (Virtual Private Networks):
Secure connection between SICOP (Integrated Public Procurement System) at the Comptroller’s Office and the Ministry of Finance, enabling access to legacy databases in the Ministry’s cloud environment.
Azure App Service con WordPress:
Development and deployment of the Observatory portal for data visualization.
ML (Machine Learning) Processes:
Execution of predictive models, such as linear regression, to optimize data analysis.
