Artificial Intelligence–Enabled Business Analytics for Strategic Project Planning and Control

Main Article Content

Md. Tamzidul Alam Chowdhury
Bipul Dutta
Mehedi Hasan
Bithika Shabnam Bithi
MD AMIT HASAN

Abstract

Strategic Project planning and project control are the key issues in multi-project environments with financial instability, resource dependencies, and time-based uncertainties. Conventional forms of project control, such as deterministic forecasting and variance-based monitoring, are largely retrospective and do not offer the ability to learn. This study suggests an AI-based predictive-prescriptive system of portfolio-level project governance fusing advanced ensemble machine learning systems with explainable and decision-support systems. Based on the Kaggle Project Portfolio Data, predictive models were created to predict the cost overrun ratio and categorise schedule delay risk. Relative analysis with traditional regression benchmarks proves substantial performance gains, and gradient boosting and stacking ensemble techniques have better predictive ability and greater power of discrimination. In a bid to be more transparent, SHAP-based explainability was added to find out the key major drivers of risks, such as budget variance and the intensity of resource utilisation. Moreover, the decision engine based on predictive risks and implemented in a prescriptive form was created to convert the forecast risk scores into an organised mitigation plan, which could then be actively addressed by means of resource redistribution, timetable buffer, and governance warning. The findings suggest that ensemble learning would help improve the precision of forecasts and interpretability, as well as prescriptive logic, which contributes to the strategic priority and robust portfolio-based decisionmaking. The suggested architecture advances information systems engineering by transforming AI-based analytics into a cohesive system.

Article Details

Section
Articles