Indonesia is one of the largest coconut producers in the world. Data and statistics from the Indonesian Ministry of Agriculture show that the coconut plantation industry in Indonesia has continued to experience significant growth in recent years. However, despite this positive growth, the coconut plantation industry in Indonesia still faces several challenges, including problems in distribution, variations in product quality, and lack of fair market access for small and medium-sized producers. To address these distribution issues, previous research illustrates the effectiveness and benefits of e-commerce in specific sectors and low-to-middle class businesses to increase sales. The use of Machine Learning and Artificial Intelligence has also been proven as a valid solution to increase sales, interact with inexperienced customers, and address quality and variety issues of fresh produce such as coconuts. From these problems, this research aims to produce an intelligent e-commerce web that integrates various intelligent features based on Artificial Intelligence. These intelligent features include the ability to detect the quality of coconut as well as providing chatbot assistance for users. The research methodology used is observation, interview and literature study. While the system development method uses the Waterfall method to produce a website with good quality and system documentation. Black-box testing and System Usability Scale (SUS) are also used to test the functionality and usability of the application. The results of this research produce E-commerce that can be used by consumers and producers to buy or sell products, and make it easier for consumers and producers to verify the quality of coconut purchased or produced. Black-box testing was conducted to verify the functionality of the system based on the specified specifications and needs. The results of black box testing showed that there were no defects or errors that may be experienced by users during the testing process. E-commerce created gets an SUS score of 63.5 which means the system is quite acceptable to users.
Kata Kunci:E-commerce;Black-box testing;System usability scale;Artificial intelligence