Elections serve as a way for the people to express their opinions about the performance of a government and can alter their future to a great extent. In the last few years, there has been an exponential increase in the number of people openly expressing their political views through platforms like Twitter. This makes Twitter one of the best sources for analysing the public pulse. The major gaps observed in existing approaches include, inability to handle tweets in multiple languages, failing to handle fake Twitter handles, and failing to discard spam tweets. In addition, very few studies were performed with respect to Indian context. To overcome these gaps, we have come up with a solution that would specifically work in the Indian context and is capable of handling tweets in Hindi in addition to English. The data related to all major political parties is collected from Kaggle and various pre-processing steps are performed. The model uses Langdetect module to segregate Hindi tweets from the corpus. Subsequently, Googletrans module for translation. Our proposed system uses the RoBERTa model for performing opinion mining on the data sets. The results obtained will show the performance of our approach.