This episode focuses on "Machine Learning Operations (MLOps)," an Independent Research and Development (IRAD) project by a team of GTRI researchers that was presented in 2023 during GTRI's IRAD Extravaganza. A noteworthy part of the IRAD Extravaganza is the IRAD of the Year Ceremony, which awards particularly outstanding projects. For each annual IRAD Extravaganza, projects are nominated for "IRAD of the Year" awards. Finalists for the IRAD of the Year were judged in two categories: Large Investment Projects, with multiyear funding greater than $50,000. Small Investment Projects, which have one-year funding of $50,000 or less.  The "Machine Learning Operations (MLOps)" project won in the Large Investment Projects category.  Research team members Maia Gatlin and Austin Ruth are the guests in this podcast episode. Gatlin and Ruth are both Research Engineers in GTRI's Electronic Systems (ELSYS) Laboratory. This IRAD focuses on the development of Infrastructure as Code (IaC) to create a deployable platform of various tools for Machine Learning Operations (MLOps). The team has successfully deployed and tested the infrastructure to showcase the benefits of the platform through various use­  cases. The primary goal is to show that the infrastructure in place can not only support inference and training of machine learning models but also can incorporate active learning and continuous delivery of models to specified repositories. With the IaC, the platform is also deployable to edge and fog machines to perform tasks at the supported resource level.

Ver todos los episodios