Background: Accurate segmentation and characterization of brain tumors from magnetic resonance imaging (MRI) are paramount for diagnosis, treatment planning, and monitoring. While deep learning models like U-Net have set a high standard for segmentation, they can fail to detect complex, multifocal disease and often lack the tools for in-depth clinical characterization beyond basic volume. Methods: A comprehensive, web-based platform built upon a hybrid Convolutional Neural Network (CNN) and Transformer architecture is presented. An end-to-end workflow is provided by the system from medical image upload (DICOM/NIfTI) to final analysis. It features a dedicated, interactive 3D reconstruction environment with real-time controls for mesh manipulation, lighting, and data export (STL for 3D printing, JSON for reports). All measurements are performed in native medical imaging units (millimeters) to ensure clinical accuracy. The platform also includes a detailed analysis tool for calculating a full suite of morphological and clinical metrics, including an estimated WHO grade. Results: Quantitatively, the Hybrid model achieved a mean Dice coefficient of 0.91 and a mean sensitivity of 0.94 across the test set, outperforming the U-Net (0.86 Dice, 0.88 sensitivity) and a traditional algorithm (0.72 Dice, 0.75 sensitivity). In a representative case of multifocal glioma, the hybrid model identified three distinct tumor foci with a total volume of 67,480 mm³, whereas the U-Net identified only a single mass of 15,140 mm³, representing a 4.4-fold increase in detected tumor burden. These results were visualized and explored in the platform's interactive 3D viewer, which provided real-time statistics and allowed for immediate export of the 3D model. Conclusion: Our work demonstrates a complete platform that not only leverages a state-of-the-art segmentation model but also provides the necessary tools for interactive visualization, analysis, and data dissemination. By seamlessly integrating a high-performance algorithm with a user-centric interface, the system serves as a powerful tool for medical education, clinical training, and reproducible research.