Impact of social media is increasing and is being embedded in daily life as it provides a platform for people to express their opinions, sentiments and feelings towards any particular topic. So the data from social media is used by researchers as a data source for opinion mining or sentiment analysis. It can be used to discover political preferences of user's regarding elections as it has become the most important tool for communication over the internet and provides with public reactions. As my research project I used Twitter data to predict the outcome of Indian General Election 2019 by analyzing sentiments from Twitter data about the candidates of two main parties BJP/NDA and Congress/UPA. The data was collected from Twitter from 1 April 2019 to 15 May 2019 using Twitter API by crawling tweets from hashtags on either party. Then preprocessing of data was performed like tokenization, lemmatization, stemming, stopword removal etc. The prediction model was created using a hybrid approach that included Lexicon-based method with Machine Learning model. The Lexicon-based approach was used to train the model and the trained model was then used to infer the sentiments along with the prediction result using Support Vector Machine. It was observed that the positive sentiment tweets were more for NDA then that of UPA, with 68% win for NDA and 32% for UPA.