The necessity of contemporary technology like the Internet of Things, the cloud, and artificial intelligence (AI) in agriculture is emphasized by smart farming systems. The digital revolution quickens traditional agricultural methods to boost crop yield while maintaining quality. Previous attempts to successfully move agricultural practitioners toward AI through sensor technology integration failed. In order to facilitate the deployment of a smart farming system with low energy consumption, we therefore suggest an architectural model consisting of four layers: sensor, network, service, and application. Furthermore, a deep learning technique is employed, concentrating on the application layer, to create a fertilizer recommendation system that is in line with the expert's judgment. For farmers' convenience, the results of the entire system are finally shown as a single mobile application. Internet of Things and Machine Learning for Predicting Pesticides and Fertilizers The Agricultural Census of India estimates that 16–17% of the nation's GDP comes from agriculture, which employs 64.5% of its workforce. Despite being the backbone of our nation, agriculture receives very little attention and either little or no development. As the world's second-largest producer of rice, our nation must adapt and concentrate on enhancing agricultural practices to enhance the quality of life for farmers. It is imperative that contemporary technologies be used in agriculture.AI gives farmers access to real-time information that helps them make informed decisions at every level of farming. This wise choice results in fewer product and chemical loss as well as cost and time savings. © 2024 Elsevier B.V., All rights reserved.