The aim of this research work is to develop a Scientific Speech Corpus of Bodo Language containing necessary information required for the design of Text To Speech Converter. In this present study, the design criteria of a speech data base called as Speech Corpus which may provides us the necessary data and information in speech research purpose are discussed thoroughly with reference to a Indian Language, Bodo. After generating the speech corpus, an attempt is made to develop an Annotation model of Bodo language by following specific policy, termed as Tagging policy. For the development of TTS (Text to speech) converter from continuous speech of a language, the role of Annotation model by following Tagging policy for that particular language is so much important. So, this annotated speech corpora can help the researchers in TTS (Text to Speech) transparency of the Bodo language through computing machine.