Sanskrit being the most culturally rich language has many deep insights and is diverse in all fields from religious texts, scientific and philosophical documents and poetic rhymes. In the modern world in today’s time the Sanskrit language has been gaining a lot of interest in the western youth as they have understood the importance of this language, so to make them accessible the Sanskrit language in easy format globally the paper discusses an approach to translate it into English. This paper highlights different approaches used for Sanskrit to English Translation and presents an idea of using Neural Machine Translation (NMT) approach for Sanskrit-to-English translation utilizing tokenization, word embedding via Word2Vec, and a Transformer-based architecture. The paper also aims to employ reinforcement learning and transfer learning techniques to enhance translation accuracy while addressing the complexities of Sanskrit’s morphology and flexible syntax. The dataset when used will be extracted from It has a large scale corpus from Sanskrit to English Translation. This corpus contains Sanskrit Shlokas and their English translations extracted by two epics viz. The Ramayana and Mahabharata. It also discusses the key challenges faced by the current translation models and highlights their issues, and further proposes an approach that aims to resolve them through a new model.