Understanding a customer’s personal opinion is extremely important to initiate and maintain a meaningful conversation. In this paper, we propose an approach to extract latent emotional beliefs of customers and use them to tailor a chatbot’s conversation. We present a machine learning based mechanism to process customer complaints and extract sentiments like customer is sad, happy, upset, etc. Further, we also train a model that extract more fine grain sentiments like the customer is irritated, harassed etc. in context of a particular complaint scenario. This information helps to tailor the dialog according to customer’s emotional state and hence improve the overall effectiveness of the dialog system.