Google plans to introduce a central repository for machine-learning tools Thursday, with the aim of making it easier for companies new to artificial intelligence to get up and running.
AI Hub will allow Google Cloud customers to access commonly used AI tools built by Google and share their own internal AI tools across their organization, at least in its alpha incarnation. A beta version of the service will open that hub up to tools developed by Google partners or “a broader array of public content,” according to the company.
If that sounds familiar, it’s because Seattle’s Algorithmia has been working on something similar for the last few years. Algorithmia has raised $12.9 million — including a Series A from Google’s AI investment arm — to build out what it calls the world’s largest marketplace for AI algorithms.
Companies that are new to artificial intelligence likely lack the technical skills needed to get up and running, which is why cloud vendors have been hot on their trails over the last several years. Google has made AI a key component of its cloud marketing strategy, hoping to cash in on the advances that its teams at Google Research and DeepMind have achieved over the past few years.
Google also announced Kubeflow Pipelines Thursday as a way for data scientists who’d prefer to roll their own to create machine-learning models. It’s an add-on to the Kubeflow open-source project Google introduced late last year.
It’s going to take a long time for the expertise required to create modern machine-learning models to spread throughout the tech industry, which is why cloud companies are so eager to offer that expertise as a service. While GitHub might have 28 million software developers on its coding service, Google estimated that there are 2 million data scientists working around the world.