Twitter is a space for online dialogue already established among the most popular social media in the world, comprising a large source of authentic data. However, many linguists feel hesitant in incorporating online data in linguistic research since online sources tend to be volatile, unpredictable, and are hardly compatible with traditional methods of analysis. Thus, in this presentation, we seek to report our experience with the software R combined with the Twitter platform in order to map the challenges and benefits faced throughout this journey. We will take into consideration the challenges of working with an unstable and untraditional corpus as well as the difficulties encountered in automatic data gathering. Many linguists, such as Blommaert (2018), have already advocated for the importance of contemplating online data in sociolinguistics as the barriers between the online-offline nexus are narrowing every day. Another important argument defended by Blommaert is the importance of investigating this type of data taking into account its specific characteristics. The way language and communities circulate in the online world is not always parallel to their circulation in the offline world. Acknowledging and reporting these differences is essential to fully understand the role of language in each context. In order to do so, we must establish a dialogue between the different faculties to guarantee a collective construction of knowledge and research development. This presentation is a collaboration between students from the Faculty of Computer Science and students from the Faculty of Letters at the Universidade Federal do Rio de Janeiro.