Showing no signs of letting up in its pursuit of public cloud leaders Amazon Web Services Inc. and Microsoft Corp., Google LLC today announced new services for hybrid cloud and on-premises computing environments aimed at making its cloud services more accessible to larger groups of users.

That’s apparent on the first day of Google’s Cloud Next conference today with the launch of a new suite of cloud services aimed at developers of container-wrapped software applications, which are one of the most popular use cases for cloud computing.

Containers give developers greater agility because they allow them to build applications just once and run them anywhere, while automating much of the infrastructure and management with the Kubernetes orchestration tool. With its new Cloud Services Platform, Google is hoping to make life easier by bringing together all of the tools developers need to get containers up and running on any platform, via a single suite of services.

The core component of Google’s Cloud Services Platform is the open-source service mesh Istio, which is a kind of framework that enables developers to connect, manage and secure microservices, or the components of their containerized applications. Istio first came available in beta back in May 2017, and after extensive testing is soon to be launched into general availability with version 1.0, meaning it’s finally ready for production deployments. The Cloud Services Platform will also offer a managed version of Istio that comes complete with additional service discovery, intelligent traffic management and secure communication enhancements.

The other main component of the Google Cloud Services Platform will be the on-premises version of its Google Kubernetes Engine service, which provides a management layer for deploying containerized applications. GKE On-Prem will soon be made available in alpha test mode, enabling developers to begin modernizing their existing applications without necessarily having to move them to the cloud, the company said.

“We want Google Cloud to be the best place to create software,” Google Cloud Chief Executive Diane Greene (pictured) said this morning.

James Kobielus, an analyst with Wikibon, owned by the same company as SiliconANGLE, said the announcement shows Google is essentially pioneering Kubernetes as the foundation of its end-to-end hybrid cloud application platform. He added that this makes sense for Google because it has a strong advantage in this area.

“It’s a bit further ahead of AWS, Microsoft and other cloud vendors in this regard, with a keen focus going forward on building up its partner ecosystem in the GCP Marketplace, of third-party cloud services apps that plug into Kubernetes running on its Cloud Services Platform,” Kobielus said.

Despite this advantage, the on-premises version of GKE may not be to everyone’s tastes, a second analyst said. Patrick Moorhead of Moor Insights & Strategy said GKE On-Prem was the most interesting of Google’s cloud announcements, but said the company still doesn’t offer a complete infrastructure stack to compare with its rivals’.

“Enterprises still need to piece-part their environments together, unlike Azure Stack, which is a full on-prem stack with a direct connect to Azure,” Moorhead said. “And ironically, AWS still manages more Kubernetes workloads than anyone else in the industry.”

But in a briefing, Google executives were keen to stress the uniqueness of their Kubernetes offerings:

“Google is the only cloud to deliver managed Kubernetes on-premises,” said Chen Goldberg, director of engineering for Kubernetes Engine and Kubernetes, Istio and Cloud Services Platform. “It looks and behaves exactly the same.” The platform also introduces new GKE Policy Management tools that allow things such as compliance to be controlled from a central location.

A cloud for all

The launch of Google Cloud Services came as Greene delivered a keynote around the theme of “Building a Cloud for Everyone,” and the company’s other announcements at Cloud Next ‘18 made it clear it doesn’t intend to leave behind anyone who could use its cloud services.

“We’re proud of being cutting-edge, but we’re also proud of having the table stakes for what enterprises need,” Greene said before introducing Target Brands Inc. as now running on Google’s cloud.

For example, the company has been aggressively tapping organizations of all shapes and sizes with its cloud-based productivity tools. G Suite, which is a rival to Microsoft’s Office 365 service, not only brings more non-information technology workers into Google’s cloud embrace, but may also in some cases help onboard growing organizations to the company’s wider range of enterprise cloud services. G Suite also gives regular employees a chance to experience some of Google’s latest technologies firsthand.

Google today infused the platform with new machine learning technology that’s designed to boost worker productivity. New features include the introduction of Smart Reply in Google Hangouts chats, and Smart Compose in Gmail, which helps users to compose email messages more quickly by predicting their responses to specific messages.

Demo of Hangout’s new Smart Reply feature

“In addition to autocompleting common phrases, Smart Compose can insert personalized information like your office or home address, so you don’t need to spend time in repetitive tasks,” Prabhakar Raghavan, Google G Suite’s vice president of apps, wrote in a blog post announcing the updates.

G Suite also gains a new feature called Grammar Suggestions, which uses a machine translation-based approach to recognize errors in Google Docs and suggest corrections on the fly.

Alan Lepofsky, principal analyst and vice president at Constellation Research Inc., said the new ML features in G Suite would be welcome at a time when many workers are struggling in the face of “information overload” at work, with too much content coming from too many people and too many channels. He said one of the main promises of AI is that it can automate repetitive and mundane tasks, and hopefully alleviate some of the extra load. However, it won’t be as easy as it first seems, he said.

“Google is not alone in this, as companies like Microsoft have added several AI features to Office365 and Cisco has added AI to their Webex portfolio,” Lepofsky said. “Mass adoption of these features requires them to be frictionless, accurate, personalized and secure.”

Sticking with G Suite, new investigation tools added to the platform’s security center can help companies keep a better handle on their data. The new tools use AI to help admins quickly identify any users within an organization who might be infected by a virus or malware, see what documents have been shared, remove access to any specific files and perform other security-related administration tasks.

Wikibon’s Kobielus said that Google is trying to position G Suite as its encroachment strategy versus Microsoft in the latter’s core enterprise knowledge worker stronghold.

“All of the G Suite enhancements being announced tomorrow are to assure that business customers that, if they wish, migrating to Google’s suite is a low-risk proposition,” he said.

Demo of G Suite’s new security investigation tool

Google is also using its cloud as a way to get more organizations to build their own AI and machine learning-powered tools and applications. With the release of its Cloud AutoML training kit that was announced earlier this year, Google is providing an easy to use drag-and-drop interface for training AI models for a range of specific use cases.

Google already offers a number of pretrained machine learning models such as its Cloud Vision API that’s used for image recognition tasks. In addition, the company also provides services such as TensorFlow and Cloud ML Engine for more experienced data scientists to build their own machine learning models from scratch.

With Cloud AutoML, Google is now targeting the middle ground in the shape of organizations that have needs which go beyond its pre-trained models, but lack the resources and skills to build their own. The first tool in this box is AutoML Vision, which extends the Cloud Vision API to recognize entirely new categories of images. Others include AutoML Natural Language, which can be used to predict custom text categories for specific domains, and AutoML Translation, which is designed for building customized translation models.

Google realized very early on in the game that AI and machine learning are key use cases for its public cloud, as enterprises need it to scale their next-generation applications, analyst Holger Mueller of Constellation Research told SiliconANGLE. Now, though, the race is all about making it easier to adopt and use AI and ML tools, which are currently only accessible to a limited number of data scientists and other specialists, he said.

“Being able to run AI and ML in an automated way is very attractive for enterprises looking to build new applications,” Mueller said. “With Google simplifying access to its speech, language and translation capabilities, it is bringing forward some of its leading AI and ML capabilities.”

Rajen Sheth, senior director of product development for Cloud AI at Google, confirmed Mueller’s assessment, saying its goal is to try to make it possible for everyone in the world to use AI and create models for their specific purposes.

“AI is really mass customization,” Sheth said. “But it’s really hard to customize deep learning. Only about 10,000 people can create machine learning models, and there are only around 2 million data scientists in the world.”

Cloud AutoML is entering beta starting today, Google said.

With reporting from Robert Hof

Photo: Robert Hof/SiliconANGLE; images: Google