Description of Session
Artificial Intelligence (AI) assisted health systems will be essential in reducing costs and supporting scaled health systems. There is, however, a difficult transition between proof of concept of such systems and deployment into an operational environment. This is due to integration difficulty, scaling challenges, and maintenance considerations. In this workshop we will demonstrate the deployment of Tensorflow server, an architecture that allows rapid deployment of Tensorflow models. Tensorflow is a state-of-the-art open-source framework for high-performance, scalable machine learning. Even though, TensorFlow is the most popular library for deep learning models for research and production, getting started with TensorFlow can be quite overwhelming for beginners and professionals alike. We will provide an end-to-end machine learning solution that can be easily adapted to specific projects. Participants will use their own laptops to develop a simple tensorflow model, deploy the model to a Tensorflow server, and query the tensorflow model for predictions through an API query. This practical session will empower participants to build their own models and server, the building blocks that allow rapid deployment of AI models into operational systems.
