This study explores a deep learning-based approach for taxonomic classification of gene sequences, leveraging convolutional neural networks (CNNs) to improve accuracy and computational efficiency. Traditional alignment-based tools, while accurate, are often computationally intensive and less scalable. The proposed method formulates taxonomic classification as a supervised learning problem and employs a 1D CNN architecture to automatically extract sequence features for classification. The model was trained and evaluated on a curated dataset derived from the SILVA database, consisting of 44,286 filtered RNA sequences from 22 bacterial genera. A comprehensive preprocessing pipeline was implemented, including sequence cleaning, one-hot encoding, strict length filtering, and stratified dataset partitioning. Regularization techniques, including L1-L2 penalties and dropout, were employed to prevent overfitting, and learning rate scheduling was used to improve convergence. The experimental results show that the model achieved a test accuracy of 97.75%, with precision, recall, and F1-scores consistently high across all genera. Performance was also compared with BLAST, revealing that while BLAST achieved slightly higher accuracy (99.8%), the proposed CNN model demonstrated significant improvements in computational efficiency. These findings suggest that deep learning offers a scalable and practical solution for high-throughput microbial classification, especially for applications that require real-time analysis.