Agriculture is being transformed by artificial intelligence and machine learning by offering adaptable algorithms that boost efficiency and productivity. Modern technology, including precision farming and digital agriculture, leverages data-intensive techniques to optimize crop management, yield prediction, and disease detection. Plant diseases, caused by pests, pathogens, and climate variability threaten agricultural productivity, making early detection vital for intervention. While various detection systems exist, some are complex and impractical for farmers, especially in developing countries. Detecting diseases early is crucial for minimizing crop damage. This paper presents an innovative plant stem disease detector leveraging deep learning and machine learning techniques. The system integrates image analysis of plant stems with comprehensive environmental condition data to enhance disease detection accuracy. The process begins with data pre-processing, plant stem images are pre-processed and each image is labelled based on plant stem disease categories, and the dataset is split into training and testing sets. A deep learning model utilizing a convolutional neural network architecture is employed to extract relevant features from images and classify whether the input stem is healthy or unhealthy. And for further classification of unhealthy stems into various categories of disease, machine learning models including Logistic Regression, Random Forest, Support Vector Machine, Naive Bayes, Gradient Boosting, XGBoost, LightGBM, K-Nearest Neighbours, and AdaBoost are evaluated. Among these, XGBoost and AdaBoost demonstrate the highest accuracy (88%). Most models achieved a moderate accuracy, with Logistic Regression, Random Forest, Gradient Boosting, and LightGBM performing similarly well (between 70% and 85%). Finally, accuracy and other performance metrics are calculated to assess how effectively each model classifies plant stem diseases.