Monitoring microbial biodiversity dynamics is crucial to understanding the causes and consequences of environmental change. The rapid development of sequencing technologies enables us to study microbes in their natural environments using metabarcoding without culturing them. Metabarcoding targets specific genetic markers, DNA barcodes, to provide taxonomic profiles for microbes. Species identification plays an important role in providing these taxonomic profiles. The most accurate method in species identification is BLAST (Basic Local Alignment Search Tool). Although BLAST has shown to be efficient, it suffers from scalability issues when dealing with millions of sequences from environmental samples. Deep learning has emerged as a successful paradigm for big data classification. We aim to build an open-source application employing deep learning that is efficient with regard to input data to decide quickly taxonomic profiles for microbes while retaining high accuracy, to trace microbial biodiversity changes in the Netherlands, and in various parts of the world.