Alzheimer's disease (AD) is a progressive neurodegenerative disorder primarily affecting older adults, characterized by neuronal loss, cognitive decline, and behavioral disturbances, ultimately resulting in a complete loss of independence. This study outlines the development and integration of machine learning models into a web-based platform aimed at providing diagnostic support for early-stage AD. The platform leverages data from the Clinical Dementia Rating Sum of Boxes (CDRSB), as provided by the Alzheimer's Disease Neuroimaging Initiative (ADNI) evaluating Control vs all the different cognitive states present in the dataset. The platform was designed following a user-centered approach and the framework of SCRUM. For AD detection, a recursive feature elimination (RFE) method was employed for feature selection, this selection was then evaluated by the Akaike information criterion (AIC) and the accuracy metric. Next an ensemble model was developed, integrating multiple machine learning algorithms, including Logistic Regression, Artificial Neural Networks (ANN), Support Vector Machines (SVM), K-Nearest Neighbors (KNN), and Nearest Centroid classifiers. The ensemble model exhibited strong performance, achieving accuracy and sensitivity metrics above 82%, demonstrating its capability to reliably identify patients at various stages of cognitive impairment. The platform, version 1.0, supports early referral to specialists through automated notifications and report generation, serving as a valuable diagnostic aid for healthcare professionals. This work underscores the potential of machine learning in advancing early detection and intervention strategies for Alzheimer's disease, offering a scalable and accessible tool to improve patient outcomes.