Home Cloud Google Announces General Availability of AI Platform Prediction; Rolls Out New Features

Google Announces General Availability of AI Platform Prediction; Rolls Out New Features


Google Cloud has announced the general availability of its AI Platform Prediction. The new service allows enterprises to build, run, and share Machine Learning (ML) models in the cloud. The AI Platform Prediction allows users to create a machine learning environment over Google Kubernetes Engine (GKE) managed services, thereby eliminating the need to set up a production-grade ML environment. Already equipped with conventional features such as auto scaling, access logs, and request/response logging, the platform has also received several new updates.  

Google also introduced new endpoints in three regions — US-central1, Europe-west4, and Asia-east1. Additionally, Google has added the feature of isolating the machine learning models to provide better security by keeping them separated from other cloud environments in the organization. The administrator can set up a perimeter where the model can only access the resources it needs. This level of security and isolation aims to provide safeguard against breaches.  

Another new feature of the new AI Prediction platform is Resource Metrics. It allows administrators to view the models’ cloud infrastructure utilization via Google’s Cloud Console and Stackdriver monitoring tools. This will help organizations to make optimal utilization of resources. 

“AI Platform makes it simple to deploy models trained using these frameworks with just a few clicks — we’ll handle the complexity of the serving infrastructure on the hardware of your choice,” Google engineers Bhupesh Chandra and Robbie Haertel reported. “With as little as one command, you can create a policy that governs existing and new VMs, ensuring proper installation and optional auto-upgrade of both agents,” explained Google Cloud product manager Morgan McLean.  

Google has made it easier to deploy models created using frameworks like XGBoost and Scikit so that organizations now have more options when it comes to running projects on AI Platform Prediction. While XGBoost can build models by a method called gradient boosting, similar to deep learning used for analyzing spreadsheets, Scikit is an AI creation tool that is far simpler to use.  

Another set of features Google announced alongside the AI Platform Prediction are monitoring features that can track the health of cloud instances. It can be installed by administrators on their instances to track metrics like memory usage. Other than this, Google also stated that the AI platform integrates with other AI technologies to simplify the workflow and improve productivity. 


Cloud Management