Cloud Next 18 - Google moves up the AI stack into enterprise apps

Profile picture for user pwainewright By Phil Wainewright July 24, 2018
Summary:
Introducing contact center and recruitment solutions today, Google Cloud signals its move up the AI stack to compete in the enterprise apps space

Google Next 18 Fei-Fei Li Cloud AI by @philww
Fei-Fei Li, Google

With the announcement of Contact Center AI and Talent Solution at Google Cloud Next today, Google is bringing its AI capabilities into the enterprise application realm. These are the first of many solutions in development that are designed to harness AI to help people work better, says the vendor.

Google is also showing off new machine learning tools and AI-powered features in G Suite — more on those announcements below — but the shift towards more packaged business solutions marks a significant expansion of Google's enterprise footprint.

Currently being tested by customers such as eBay, Contact Center AI combines with partners in the contact center space such as Genesys to provide an end-to-end offering that supports human agents with AI-powered capabilities. It means that a customer call can be answered by a virtual agent before being passed over to a contact center agent, who is then supported by having relevant information and actions suggested by AI services that monitor the conversation. The solution builds on a development suite for building conversational agents called Dialogflow Enterprise Edition, which Google introduced late last year.

Building human-centered AI into apps

It's all part of a mission to build what Google calls "human-centered AI," says Fei-Fei Li, Chief Scientist at Google Cloud AI:

Context sensitive AI elevates human talent ...Rather than replacing human skills, AI's greatest potential lies in enhancing humanity.

This is the "first of many" such solutions, she adds in a blog post published today, including a recruitment app that targets the HCM space:

Our Talent Solution is built on top of deep learning matching technology to improve enterprise talent acquisition — saving recruiters time by finding the best past applicants for a role, and helping candidates find the right jobs that match their skills and interests.

Today also saw further expansion of the AI and ML features in the G Suite collaboration platform. These include suggested auto-completion of text when writing emails — a feature that's already in Gmail for consumers, and is now rolling out to G Suite — extension of the Smart Reply function that's already in Gmail to Hangouts Chat, and an early adopter program for AI-powered grammar suggestions. Google is also bringing its Voice Assistant technology into the business arena, with the introduction of voice commands to its Hangouts meeting room system. Another feature being introduced is a facility to automatically suggest suitable room bookings when scheduling meetings.

Build your own AI models with AutoML

Among AI services, Google added several new capabilities to its AutoML function, which uses machine learning to automatically build ML models from custom datasets. For example, AutoML Vision, now in public beta, can be used to recognize new categories of images. Fashion retailer Urban Outfitters found that it was faster and more accurate at building a model for recognizing fashion images than its own in-house ML systems, according to Li. One advantage of AutoML over predefined models, she adds, is that the resulting model remains proprietary:

Your data is not only private but should remain your competiive advantage. None of the data that trains an AutoML model is used in any other Google machine learning application.

In addition to the image tool, AutoML is now available in beta for natural language and translation.

Li also emphasizes Google's commitment to ensuring the ethical use of AI:

Algorithmic bias is a concern for everyone. We've invested heavily in a set of best practices to help you build models that you can trust to treat your users with fairness.

Other AL-related announcements today include the launch of the third generation of Google's TPU machine learning co-processors, which customers use for high-speed processing of their own ML models.

G Suite is today available in data regions at no additional cost. This allows customers to set up multi-region policies for data management — useful in the wake of the introduction of GDPR and similar data protection regulations. On the security front, the recently launched security center in G Suite is now gaining an investigation tool which helps administrators identify and take action on potential security threats, such as abnormal sharing of files.

My take

AI's been seen as one of Google's strengths as a cloud platform, but the decision to extend into the application layer with packaged solutions signals a new direction, and one that puts it into direct competition with the likes of Salesforce and Workday. It's a measure of how far the Google Cloud business has come since Diane Greene took the helm as CEO, but also of the extent to which AI is changing the enterprise computing landscape.