Cloud acceleration of INSTINCT

The advent of Artificial Intelligence/Machine Learning (AI/ML) in marine mammal science has brought about new challenges in development and support of these applications. As scientists journey into the world of software development, successful approaches developed in-house face the challenge of acceleration to more powerful computing infrastructure. The AI/ML in-house application INSTINCT (the Infrastructure for Noise and Soundscape Tolerant Investigation of Nonspecific Call Types) is an autodetection system that identifies marine mammal call types of interest in large acoustic datasets. In this report, INSTINCT was reengineered to allow acceleration to hyperscale in a cloud production environment. This deployment pattern was assessed experimentally, and recommendations based on our findings are included.