The study of environmental DNA using next-generation sequencing (NGS) has yielded important new biological insights, revealing animals and plants living on Greenland 2 million years ago and monitoring present-day species in the ocean. Sequencing of environmental RNA (eRNA) can yield new information beyond the presence of species, including: organismal developmental stage and age; health, stress and disease conditions and the presence of RNA viruses. Recent studies have demonstrated that RNA fragments can persist outside of cells for far longer than previously assumed, and that these fragments can be detected by next-generation sequencing. Yet, the computational tools to leverage the specifics of eRNA data do not exist, hampering the field. We will develop an integrated framework to trace eRNA sequences to their species of origin, combining Kraken2 with dedicated analyses for abundant and informative microRNA and rRNA sequences. We will further train and test a Deep Learning Categorical Classification algorithm to deconvolute these measurements into exact estimates of the contributing species and their relative molecular contribution as a percentage. Our software - Traident - will for the first time allow researchers to analyze eRNA data, fully leveraging on the information captured in abundant transcripts such as microRNAs and rRNAs - with potential implications for research on conservation efforts, museomics, ancient RNA and food safety.