<title>Abstract</title> Insufficient water, below 70%, limits cocoa growth, reduces production and affects quality due to water stress. On the other hand, excess moisture, above 85%, obstructs air channels in the soil and causes root rot, reducing nutrient absorption and crop yield. These unfavorable water conditions negatively impact both cocoa quantity and quality. Accurate irrigation management, staying within an optimal range of 70-85%, is essential to maximize cocoa production and quality. The project proposes a smart irrigation system for cocoa using Edge Impulse and artificial intelligence to analyze air temperature and humidity, as well as predict rainfall with cloud data. Sensors measure soil moisture in real time near each plant. The system compares the rain prediction with soil moisture, triggering drip irrigation only when moisture is predicted to be lacking and rain is not expected. This integration of Edge Impulse improves efficiency and provides high-performance real-time data analysis. A smart irrigation system was implemented, combining data from DTH11 sensors and hygrometer with Edge Impulse, and the algorithm was transferred to an Arduino Uno to control the drip irrigation motor pump in real time. Irrigation is activated only under optimal conditions, considering air and soil moisture. This efficient approach reduces water consumption and optimizes the energy used in irrigation. Integration with Raspberry Pi and Firebase for remote control establishes a scalable and sustainable model for precision agriculture, highlighting the effectiveness of Edge Impulse in modern agricultural resource management.