The agriculture sector is India’s primary source of national income and occupation. Today India is considered a global agricultural powerhouse, but the agriculture sector suffers from serious maladies. As a result, Indian agriculturists and farmers are poor. As per the experts, one of the major factors contributing to farmers' poverty is poor access to reliable and timely market information and the unavailability of commodity price forecasts. In this study, our main goal is to build a web application for Indian farmers with a simple user interface to provide them easy access to current and historical prices of their commodities as well as predict daily, weekly and monthly commodity prices using a long-short term memory (LSTM) machine learning model. For the price forecast, we evaluated different methods based on input data type and Mean Square Error (MSE). Upon evaluation, we used the LSTM model for its efficacy and precision on time-series data and lowest MSE among contemporaries.