Google Cloud’s Vertex AI enters GA for easier and faster ML model creation

Abner Li | May 18 2021 - 12:45 pm PT

All parts of Google were on display at I/O 2021 today, with Google Cloud announcing general availability (GA) of Vertex AI. The platform is aimed at making machine learning more readily available to both developers and businesses.

With Vertex AI, Google Cloud touts needing nearly 80% fewer lines of code to train a model versus competing offerings. The goal is to make building, training, deploying, and maintaining machine learning models available to “all levels of expertise.”

Vertex AI brings together Google Cloud’s various tools for building ML “under one unified UI and API.” It’s an alternative to “manually piecing together ML point solutions.”

In this single environment, customers can move models from experimentation to production faster, more efficiently discover patterns and anomalies, make better predictions and decisions, and generally be more agile in the face of shifting market dynamics.

Key features include Vertex Vizier to increase the rate of experimentation, while Vertex Experiments uses faster model selection to speed up development. Vertex Feature Store allows for easy ML capability sharing and reuse. Once released, Vertex offers tools to manage the deployment.

Google says the product incorporates what it has learned about bringing ML models to production and will benefit from continuous research updates.