PLAVIDA, PLatform for Audio and VIdeo Data Annotation is a platform designed to facilitate audio and video data annotation. To perform sound classification tasks with Machine Learning algorithms, we need annotated data on these sounds. It is on the basis of this annotated data that these algorithms will learn to make classifications. However, the community lacks labelled audio data on African languages. PLAVIDA will allow researchers the opportunity to create a multimedia labelled databases which can be used as input in Artificial Intelligence models. This could boost research around audio classification in several African languages. We have used python and Android IONIC/Angular technology to develop this tool. The innovation in PLAVIDA, is the possibility given to illiterate people to be able to interact with, when we want to labelle sound or video in African local languages. The tool can be then used both by literate and illiterate people. The type of labelling we are faced on concern the emotional perception people can have when listening or watching a media. It incorporates an annotation logic based primarily on the maximum rate of the same emotional perception over all. In the case where there is no majority vote, the user profile criterion is used. The data annotated using this application can be exported in XML, CSV or JSON format. These types of format are the data formats used to create Artificial Intelligence models.