The goal of rainfall prediction is to pinpoint the exact spot where it will rain. It is seen as essential for other businesses as well as the agricultural one. Various frequency distribution functions have various distributions of daily rainfall levels around the world. In countries with an agricultural economy, rainfall becomes important. For the prediction of the cumulative monthly rainfall, we offer a machine learning-based architecture. In this work, daily rainfall was predicted using 65 Years of Weather Data for Bangladesh (1948–2013) using polynomial regression, decision tree models, K-NN, SVM, RF models, AdaBoost regressors, stacking regressors, and ANN algorithms. R2 and RMSE test results were used to compare the projected outcomes from the chosen algorithms with the actual data in order to assess the accuracy of the predictions. We discovered that certain algorithms could estimate daily rainfall rather accurately. With these models, rainfall over the course of a year, a month, and a day may be forecast.