Google unveils new Cloud Machine Learning APIs, tools, and services
Google Cloud on Tuesday announced new offerings in Cloud Machine Learning, one of its fastest growing product areas. The new tools and services are making machine learning more accessible and giving businesses new ways to leverage the technology.
First, Google is rolling out an entirely new API to help with job searches. It illustrates how machine learning can provide a simple solution for a complicated task: finding the best job listings based on a potential employee's preferences, including their search terms and factors like seniority and location.
Early adopters of the Cloud Jobs API include CareerBuilder and Dice.com, two leading job sites. CareerBuilder built a prototype with the Cloud Jobs API in just 48 hours, and after testing it out, plans to leverage the API for its customers in the near future. Cloud Jobs API is now available in a limited alpha. Google is initially targeting the US and Canada, given that cultural factors influence the way jobs are listed.
Next, Google is lowering the price of its Cloud Vision API by as much as about 80 percent. The price reduction comes after Google started using its custom TPUs to improve performance and efficiency. The API also has significantly improved image recognition capabilities. It's now capable of identifying new entities like landmarks and can identify millions of logos. Globally available since May, the Vision API is already one of Google's most popular APIs in the cloud.
Meanwhile, Google is rolling out a premium version of its Cloud Translation API. Earlier this year, Google unveiled its Google Neural Machine Translation system, which uses advanced training techniques and runs on TPUs to reduce translation errors by as much as 55 percent to 85 percent on several major language pairs. Now, this system is available as the Translation API premium offering. For now, the API offers support for up to eight languages (English to Chinese, French, German, Japanese, Korean, Portuguese, Spanish, and Turkish) and 16 languages pairs.
The standard addition of the API, used for more than 90 different languages, is already very accurate. However, certain industries like the travel industry care deeply about getting the best possible translation.
Google is also making its Cloud Natural Language API generally available. Based on feedback from beta testers like Evernote, the API comes with improved features such as more granular sentiment analysis and improved syntax analysis.
Lastly, Google is giving businesses using Google Compute Engine or Cloud Machine Learning the option of using GPUs. Making GPUs available in Google Cloud means you can focus on solving challenging computational problems while accessing GPU machines from anywhere and only paying for what you need. The option will benefit Google customers who want to speed machine learning training time.
Google on Tuesday also announced other efforts to push into the fields of machine learning and artificial intelligence. The company announced that it's creating a Cloud Machine Learning group to focus exclusively on cloud-based machine learning solutions for enterprise. The group will be led by Fei-Fei Li, former director of the Stanford Artificial Intelligence Lab and the Stanford Vision Lab, and Jia Li, former head of research at Snapchat.
Google also unveiled AI Experiments, a set of machine learning demos built in collaboration with musicians, scientists, and others.