Abstract: Fast, accessible, and accurate identification of pests is essential in the forestry industry to ensure appropriate action is taken as soon as possible. For this purpose, the Forestry and Agricultural Biotechnology Institute (FABI) at the University of Pretoria have launched an initiative to train and deploy machine learning models capable of identifying pests underrepresented in existing models and data sets, such as the Gonipterus and Sirex pests prevalent in Southern Africa. As part of this project, we work through the data science project lifecycle to plan, visualise, build, and deploy a model capable of detecting these pests. Transfer learning is utilised by training a backbone model on the 19 000 box annotated images in the IP102 public dataset. A small dataset provided by FABI containing photos of the Gonipterus and Sirex is then box annotated for various features, and a range of models are trained to identify not only the pests themselves but also specific features of the insects. This allows the process to output predictions regarding these specific features even if it cannot predict the insect’s species, aiding researchers in the identification process. Finally, the next steps are discussed for the project partners to improve and deploy the model to a production environment.