Breast cancer remains a pressing health concern in Tanzania, characterized by rising incidence rates and consequential mortality. Traditional diagnostic methodologies, encompassing mammography, ultrasound, and biopsy, exhibit limitations concerning accuracy, subjectivity, and accessibility. In recent years, the advent of machine learning techniques, notably Convolutional Neural Networks (CNNs), has offered a promising avenue for breast cancer detection, addressing the deficiencies inherent in conventional approaches. This paper centers on the development of an AI-driven mobile application (Mobile App) integrated with breast cancer prediction model tailored for breast cancer identification in Tanzania, leveraging the capabilities of CNN models. The app will allow radiologists to capture breast MRI images through a mobile phone camera and receive predictions categorizing the captured images as Malignant or Benign, aiding in prompt diagnosis and improve the accuracy of the diagnosis. The developed model uses 30 MRI breast images from Muhimbili National Hospital (MNH). Subsequently, data augmentation techniques were implemented, bolstering the dataset to 1419 images, inclusive of 700 benign and 719 malignant cases. The resultant CNN model developed demonstrated an exceptional accuracy of 96.69%, underscoring its potential effectiveness in discerning breast cancer and its prospects for facilitating early detection within the Tanzanian context.