Google Cloud Launches Vertex AI to Streamline Machine Learning Deployment
Google Cloud introduces Vertex AI, a unified managed platform that replaces several legacy AI services to accelerate the deployment and maintenance of machine learning models.
Google Cloud announces the general availability of Vertex AI, a managed machine learning platform designed to accelerate the deployment and maintenance of artificial intelligence models. Engineers use this new service to manage various datasets and build machine learning pipelines to train and evaluate models using either Google Cloud algorithms or custom training code.
The platform provides Docker images that developers run for serving predictions from trained model artifacts, offering prebuilt containers for popular frameworks like TensorFlow, XGBoost, and Scikit-learn. For scenarios where data needs to stay on-site or on a device, the experimental Vertex ML Edge Manager deploys and monitors models directly at the edge.
Vertex AI replaces multiple legacy Google services, including AI Platform Data Labeling, AI Platform Training and Prediction, and several AutoML products. Google executives state that the goal of this unified platform is to help data scientists escape the orchestration weeds and move AI projects out of pilot purgatory and into full-scale production.