AI Platform supports Kubeflow, which lets you build portable ML pipelines. That you can run on-premises or on Google Cloud without significant code changes. As you deploy your AI applications to production. you’ll have access to Google AI technology like TensorFlow, TPUs, and TFX tools.
Machine learning development: the end-to-end cycle
Through storing your data in Cloud Storage or BigQuery. you can use the built-in data labeling service to label your training data. and applying classification, object detection, and entity extraction, etc. You can also import the labeled data to AutoML and train a model.
Build and Run
You can build your ML applications on GCP with a managed Jupyter Notebook. that provides configured environments for different ML frameworks.
You can manage your models, experiments, and end-to-end workflows using the AI Platform interface within the GCP console.