This paper is developed to take the edge off depression. Chatbots are special agents which are used to process a specific task, and it can be used to introduce a product to a customer or solve relative problems associated with a product, thus saving human resources. In this study, we have aimed to provide a chatbot for a society that will help to reduce the number of depression survivors. In this study, we have used RNN LSTM encoder-decoder model to know the user’s emotional state and according to that chatbot gives the best response. We have proposed a multi-purpose dialogue model which can be used in daily communication rather than for specific tasks. It helps those people who are suffering from depression and have fear of sharing their feelings or fear of being judged.