<p>Every smart application was improved based on today's advancements. Considering this advancement, improving the smart facilities further in the agriculture sector is the most required role. This might help the farmers create a green society by improving crop yields. Hence, an interactive voiced interface system has been introduced for this current work. It is flexible for both audio and image input data, but the output has remained in audio format. Moreover, a novel system known as Zebra MobileNet Encoder Strategy (ZMES) is implemented to attain the interactive voiced interface. Here, the executed system is tested with both audio and image input, and the prime process, like data initialization, filtering, segmentation or audio analysis, and disease prediction with a recommended solution in audio format, is formulated under the ZMES. To validate this system, the MATLAB tool was adopted for this study, and the robustness was measured with other traditional models in the form of different error parameters. In that, the newly designed ZMES scored the lowest error rate with high accuracy in disease prediction and recommendation.</p>