AI4Govern is a hackathon project developed for the NIRU AI Hackathon, focused on building an intelligent dashboard that flags potential governance and financial risks in public projects. The MVP leverages historical public contract award data from Kenya to identify high-risk contracts based on value concentration and supplier patterns.
# AI4-GOVERN
AI4Govern is a hackathon project developed for the NIRU AI Hackathon, focused on building an intelligent dashboard that flags potential governance and financial risks in public projects. The MVP leverages historical public contract award data from Kenya to identify high-risk contracts based on value concentration and supplier patterns.
# AI4Govern – Intelligent Dashboard for Public Project Risk Analysis
## Overview
AI4Govern is an AI-driven governance analytics project developed as part of the **NIRU AI Hackathon**. The project focuses on demonstrating how data science and machine learning can be used to enhance **transparency, accountability, and early risk detection** in public sector projects.
The MVP analyzes historical public procurement and contract award data from Kenya to identify potential governance and financial risk patterns, and presents these insights through an interactive dashboard.
## Problem Statement
Public sector projects often face challenges such as cost overruns, delays, and limited transparency. Oversight bodies typically rely on manual reviews and retrospective audits, which can delay intervention.
There is a need for **data-driven tools** that can flag high-risk projects early, enabling better prioritization, monitoring, and accountability.
## Project Objective
The objective of AI4Govern is to:
* Analyze historical public contract award data
* Identify risk signals associated with large and repetitive procurement contracts
* Classify projects into risk levels (Low, Medium, High)
* Provide a simple, explainable dashboard to support oversight and decision-making
## Data Source
The MVP uses historical **World Bank Contract Awards – Investment Project Financing** data for Kenya.
Key characteristics of the dataset:
* Fiscal Years: 2020 onwards
* Records: Thousands of contract awards
* Core variables include contract value, supplier, procurement method, sector, and award date
Note: The MVP relies on historical d …