Predicting the amount of rain is crucial to people's daily lives. Since the current technologies cannot accurately estimate when it will rain, many different types of individuals have been experiencing inconvenience. Commencing with the farmers who suffer the most, their crops are harmed by intense and unpredictable rainfall. Accurate forecasts are also necessary for city dwellers who commute to work in order to organize their schedules, modes of transportation, and numerous other daily activities. Thus, there is an urgent need for an early warning system that can precisely forecast when it will rain. Our aim is to use machine learning (ML) and Internet of Things (IOT) algorithms to build a system that can accurately predict rainfall. This consist of a microcontroller (Arduino UNO) which records atmosphere parameters with help of three sensors namely DHT11, MQ2 and rain sensor while they are working on the field. These values are logged into THINGSPEAK via the internet with help of a Wi-Fi module called ESP8266. These values are recorded at different instances throughout the day and are fed into the machine learning algorithms. The data is collected and pre-processed to train machine learning models, specifically Support Vector Machine (SVC), XGBoost Classifier, and Logistic Regression, to predict short-term rainfall events. The system aims to compare the machine learning techniques in terms of their accuracy of prediction, with XGBoost surpassing the other two algorithms with an accuracy of 99%.