In recent years, there have been significant developments in open source machine learning toolkits, such as scikit-learn, TensorFlow, PyTorch, and many others. These toolkits have been widely adopted by business and developers looking to introduce machine learning into their organizations. However, to deliver on their potential, they must to be embedded in an end-to-end machine learning pipeline.
Steve Flinter will outline the work that Mastercard Labs undertook to build an end-to-end machine learning pipeline, suitable for both R&D and production, using Kubernetes and Kubeflow. He demonstrates how the pipeline can be defined in software, configured, connected to a data streaming service (Apache Kafka), and used to train and deploy a model, which can be exposed for inference via an API. If you’re a professional software engineer or machine learning engineer seeking to understand how to introduce a robust ML pipeline into your organization, this talk is ideal for you. You’ll also learn how to manage the end-to-end workflow of data to training to deploying and serving a trained model.