The rapid advancement of Natural Language Processing has brought forth numerous text prediction models that significantly enhance typing efficiency. However, many indigenous languages, like Bodo, face challenges in leveraging these technologies due to the scarcity of digital resources and tools. This research focuses on developing a Word Prediction system specifically tailored for the Bodo language, addressing the lack of such tools for its speakers. Leveraging modern machine learning techniques, we propose a system that predicts the next word based on the context of the preceding text, improving typing efficiency and accuracy for Bodo speakers. The WP system is integrated into an Android keyboard application, providing Bodo speakers with real-time predictive text features to facilitate digital communication in their native language. The dataset for the model was compiled from available online resources in Bodo and preprocessed to improve its quality. This work contributes to the preservation of Bodo by making it more accessible in digital spaces, encouraging the creation and sharing of Bodo-language content, and addressing the digital divide for minority languages.