This study focuses on the early diagnosis of dengue through the analysis of blood chemistry parameters, using data collected from patients in the city of Tena. This research arises from the need for faster and more accessible diagnostic methods, especially in endemic areas where dengue poses a significant public health risk. To achieve this goal, artificial intelligence techniques were implemented, focusing primarily on dimensionality reduction to manage the large number of variables in blood data. Algorithms such as Principal Component Analysis (PCA), Incremental PCA (IPCA), and Kernel PCA (KPCA) were used to identify the most relevant features and reduce model complexity. Additionally, ensemble methods like Bagging and Boosting were employed, which combine multiple predictive models to optimize system accuracy. The preliminary results obtained are encouraging. The use of ensemble algorithms, particularly those based on Boosting, showed superior performance in detecting dengue. Specifically, the Support Vector Classifier (SVC) achieved the highest score under the Bagging scheme, with a value of 0.6928, suggesting that this approach offers a more accurate and timely diagnosis compared to other methods. This data is processed through an API that generates a result indicating the possible presence or absence of dengue. This technological integration not only accelerates the diagnostic process but also facilitates access to the tool in different clinical contexts. The results obtained suggest that this AI-based approach has great potential to improve early detection of dengue in resource-limited settings, offering a more viable and accessible alternative compared to conventional diagnostic tests, thereby contributing to more efficient public health management.