Category Archives: Google for Work Blog

Work is going Google

Say hello to the new Hire, a smarter, faster way to recruit, by Google

With thelaunch of Hire last year, we simplified the hiring process by integrating it into the tools where recruiters already spend much of their day—Gmail, Google Calendar and other G Suite apps. Recruiters tell us Hire has fundamentally improved how they work, with less context switching between apps. In fact, when we measured user activity, we found Hire reduced time spent completing common recruiting tasks—like reviewing applications or scheduling interviews—by up to 84 percent. But we wanted to do more.

The result is our latest release of Hire. By incorporating Google AI, Hire now reduces repetitive, time-consuming tasks, into one-click interactions. This means hiring teams can spend less time with logistics and more time connecting with people.

Here’s a little more on what recruiters can do with the new Hire:

Schedule interviews in seconds

Recruiters and recruiting coordinators spend a lot of time managing interview logistics—finding available time on calendars, booking rooms, and pulling together the right information to prep interviewers. To streamline this process, Hire now uses AI to automatically suggest interviewers and ideal time slots, reducing interview scheduling to a few clicks.

G Suite Hire - Replace Interviewer

If an interviewer cancels last minute, Hire not only alerts you, it also recommends available replacement interviewers and makes it easy to quickly invite them. This means hiring teams can invest time in preparing for interviews and building relationships with candidates instead of scheduling rooms and checking calendars.

Auto-highlight resumes

A huge portion of recruiters' time is spent reviewing resumes. Watching people interact with Hire, we found that they were frequently using “Ctrl+F” to search for the right skills as they scanned through a resume—a repetitive, manual task that could easily be automated. Using AI, Hire now automatically analyzes the terms in a job description or search query and auto-highlights them on resumes, including synonyms and acronyms.

G Suite Hire - Auto-highlight Resumes

Click to call candidates

Whether they’re screening candidates, conducting interviews, or following-up on offers, recruiters often have dozens of phone conversations each day. This means spending a lot of time searching for phone numbers or logging notes. Hire now simplifies every phone conversation with click-to-call functionality, and automatically logs calls so team members know who has spoken with a candidate.

“Using Hire by Google has helped streamline the processes that used to take up a lot of my time, which allows me to focus on the next steps to make sure candidates have the best experience.” – Anna McMurray, Chocolate Talent Scout, Dandelion Chocolate


There’s a huge opportunity for technology—and AI specifically—to help people work faster and therefore focus on uniquely human activities. Ultimately, that’s what Hire is all about, and the functionality we’re adding today demonstrates our commitment to help companies focus on people and build their best teams.  Visit hire.google.com to learn more.

Source: Google Cloud


Build the apps your business needs with App Maker

It’s easy to get used to doing things a certain way. This can be a good thing if you're prepping for the World Cup, where practice, repetition and routine can make you a champion. But if you're like the rest of us who rely on certain processes to get things done at work—like budgeting or filing expenses—“trusting the process” doesn’t always feel rewarding.

Today, we’re making App Maker generally available to help you rethink how your teams operate. App Maker is G Suite’s low-code application development environment that makes it easy for teams to build custom apps to speed up workflows and make processes better.

Apps to fill business gaps, built for your needs

Analysts estimate that the right custom mobile app can save each employee 7.5 hours per week (that’s a week’s worth of lunch breaks!). Yet, too few businesses have the means, let alone the resources, to invest time and effort in building custom apps. Why? Because their IT budget centers on big enterprise apps like CRM, ERP and SCM and beyond those priorities, IT executives’ attention focuses on security and governance.

App Maker was created to enable your line-of-business teams to build apps for the jobs these bigger apps don't tackle. With App Maker, you can revamp company processes like requesting purchase orders or filing and resolving help desk tickets, as if you designed and built the processes yourself.

Pooling talent resources was always an ad hoc process, but App Maker let us quickly build an app that tracks allocation requests in detail. Peter McAuley
Director of IT, EA

Take Electronic Arts (EA) for example, a global interactive entertainment software company that makes games and more. To create many of its games, EA has to occasionally allocate staff resources to different projects—whether it’s a designer needed on a new game project or an HR consultant to advise on setting up a studio. EA’s IT department used App Maker to create a custom app to streamline capital resourcing.

“Pooling talent resources was always an ad hoc process, but App Maker let us quickly build an app that tracks allocation requests in detail,” says Peter McAuley, director of IT at EA. “Our custom app also calculates and provides management with a view of total resource utilization by month, something which was always more of a chore to put together manually.”

Over the last few months, we've worked closely with customers and partners around the world, including EA, Colgate-Palmolive, SADA Systems and more, to build apps to solve specific business needs.

Create custom apps in G Suite with App Maker

Powerful new database model and governance

Since launching App Maker in our Early Adopter Program, we've made changes and added new features and improvements to make it even more useful for our customers.

  • Open:App Maker now offers built-in support for Cloud SQL (GCP account required), offering high performance, scalability and convenience. It also supports a Bring Your Own Database (“BYODB”) model, letting you connect it to your own database using JDBC or a REST API.
  • Fast: Responsive templates, samples, a drag-and-drop UI design and declarative data modeling make it easier for IT developers, or anyone who wants to automate work, to design and build apps even faster.

App Maker
  • Connected:Whether it’s Gmail, Calendar or Sheets, App Maker makes it a cinch to connect with the data and services you need, all while making your apps more powerful. Plus, you can use Apps Script to accessover 40 Google services, Google Cloud Platform and other third-party services that support JDBC and REST.
  • Managed: G Suite administrators now have visibility over the apps running in their organization including owners, usage metrics and OAuth permissions. Expanded OAuth Whitelisting controls mean administrators will also be able to prevent apps from running without their approval. Learn more.

Get started

App Maker is now available to all G Suite Business and Enterprise customers, as well as G Suite for Education customers. Learn more, or see how you can get started with documentation and this codelab. We can’t wait to see what your companies build with it.

Source: Google Cloud


Bring teams together with new G Suite integrations

The average employee uses 36 cloud services at work to collaborate or share files. That’s more apps than hours in a day! With so many apps and services, it’s important that your business  chooses modern, open tools that make it easy to work with anyone—be it inside or outside of your organization.

We want to reduce complexity for your teams, which is why we’ve built integrations in G Suite like third-party add-ons in Gmail and Google Docs and better interoperability with Microsoft. Today, we’re announcing new integrations in Hangouts Meet and Calendar to help you have even better meeting experiences regardless of the conference technology you use. We’re also bringing you new ways to collaborate in Hangouts Chat and Sheets.

Collaborate easily with new compatibility in Hangouts Meet

Nothing’s more frustrating than hosting a meeting and having trouble getting people to join because of issues with technology—it can interrupt workflows and slow down productivity. We want to make it easier for businesses to use meeting solutions, like Hangouts Meet, without worrying about compatibility with existing equipment. This is why we’re making Meet compatible with traditional video conferencing systems, like Polycom and Cisco. In partnership with Pexip, teams will be able to join a meeting on Meet from their preferred equipment in the coming weeks. We’re also making it possible for Microsoft Skype for Business users to easily join a meeting on Meet directly from their Skype app.

Spotify has benefited from using Hangouts Meet to collaborate both internally and externally. Says Martin Antonsson, AV Infrastructure Engineer at Spotify, “Interoperability between Hangouts Meet and other video conferencing systems is fundamental to enabling collaboration. Now we can focus on having productive meetings instead of worrying about the technology.”

Now we can focus on having productive meetings instead of worrying about the technology. Martin Antonsson
AV Infrastructure Engineer, Spotify

Build add-ons in Google Calendar, join meetings in a cinch

A big part of having productive meetings is making sure that the right folks are able to join them. That’s why we’re announcing support for third-party conferencing natively in Google Calendar. Soon conference providers can easily build add-ons to create, view and join a video conference directly from a Calendar event. Users can simply click a link in the Calendar invite to join on web or mobile.

Interop Image 1

Cisco Webex is building an add-on to help users easily schedule meetings right from Google Calendar without requiring a download or plug-in. Arkadin, GoToMeeting, LogMeIn, Dialpad, RingCentral, Vidyo and Vonage are also building add-ons to help them more easily meet with their customers. These third-party conferencing add-ons will be available in the G Suite Marketplace in the coming months with details to come for developers.

You might have seen that we made it easier to view schedule availability across Google Calendar and Microsoft Exchange last year. Building on this, we’re also announcing a way for you to book resources like rooms, equipment and more in Microsoft Exchange. If you use G Suite, you’ll be able to easily view and book resources stored in Exchange and Office 365 in the coming months.

Work with teams outside of your domain in Hangouts Chat

After you get the right people in the room with the right setup, it’s important to be able to share information in real time during your meetings—even with folks who may not be in your company. In the coming months, you’ll be able to include people from outside of your organization in Chat, making it easy to stay aligned with clients, vendors, partners and others, all from one place.

Resuelve, a consumer debt management business based in Mexico, has been using Chat to collaborate within their organization, particularly as they look to grow their presence in additional regions. "Hangouts Chat has been crucial to our company's ability to expand into other markets,” says Jordi Adame, Chief Technology Officer of Resuelve. “It’s helped our internal teams be productive and we're looking forward to connecting with people outside of our organization in a similarly efficient way.”

New SAP integrations with Google Sheets

Companies often have critical business data in their SAP systems. In an upcoming SAP release, employees will be able to discover additional insights from their ERP content by importing it directly into Google Sheets. With this new integration, you can skip manually exporting data to CSVs and uploading them to Drive. Instead, export directly to Sheets and analyze data with tools like intelligent pivot tables. You can also skip tedious formatting by recording macros in Sheets, making it easier to streamline business processes and share information across teams right away.
SAP Sheets integration

Looking ahead

In the coming months, you’ll be able to join conferences on Hangouts Meet from your existing meeting room hardware, book rooms from Microsoft Exchange in Google Calendar and collaborate with folks outside of your domain in Hangouts Chat. You’ll also start to see integrations with Google Sheets in an upcoming SAP release, too.

Learn more about how your business’ technology can co-exist with G Suite. Visit the Next ‘18 website to register.

Source: Google Cloud


New collaboration with Fitbit to drive positive health outcomes

As I’ve shared in the past, Google Cloud’s vision for the healthcare industry is very much a reflection of Google’s overall mission. We’re building healthcare-specific products and solutions as well as supporting a growing partner ecosystem to help companies organize healthcare data in a way that is accessible and interoperable, but also secure, enabling them to create a positive and lasting impact on human health.


Today we’re announcing a new collaboration with Fitbit across wearables and digital health that will help drive positive health outcomes at scale. Fitbit has chosen Google Cloud as their preferred cloud provider and will be using our Cloud Healthcare API to provide an interoperability solution that enables their users to collaborate on care with their own healthcare providers. For example, using Cloud Healthcare API with Fitbit’s newly acquired Twine Health platform. The two companies will also be exploring how Google Cloud’s machine learning APIs can help them uncover deeper insights to benefit their users. 


Although we’re just getting started in this collaboration, we’re excited by what’s possible. To date, Fitbit has sold more than 76 million devices, built a community of more than 25 million active users and has one of the world’s largest health and fitness databases. We hope that by helping them take advantage of our highly secure platform and support for open standards, we can bring better health to more people around the world.


Source: Google Cloud


Google Cloud Platform announces new credits program for researchers

From Big Data projects like Strayer University’s student support system to AI projects like Carnegie Mellon’s socially aware robot, researchers are discovering that cloud technology can help make academic research cheaper, faster, easier, and more secure. Whether you’re just starting out with a new idea, or validating your work before sharing it with the public, we want to help you advance your new discoveries. That’s why we’re deepening our support for your biggest questions and best guesses through a new program: Google Cloud Platform (GCP) research credits. Academic researchers in qualified regions are encouraged to apply.


Like the Google Cloud Platform Education Grants to support computer science courses and the partnership to support National Science Foundation (NSF) grants in BIGDATA, our GCP research credits program supports faculty who want to take advantage of GCP’s data storage, analytics, and machine-learning capabilities. Andrew V. Sutherland, a computational number theorist and Principal Research Scientist at the Massachusetts Institute of Technology, is one of a growing number of academic researchers who have already made the transition and benefited from GCP. His team moved the L-Functions and Modular Forms Database to GCP because “we are mathematicians who want to focus on our research, and not have to worry about hardware failures or scaling issues with the website.”

Other researchers are taking advantage of GCP’s scalable infrastructure. Ryan Abernathey, Assistant Professor of Earth and Environmental Sciences, Ocean and Climate Physics at the Lamont-Doherty Earth Observatory at Columbia University, used Google Cloud credits through an NSF partnership and, with his team, developed an open-source platform to manage the complex data sets of climate science. The platform, called Pangeo, can run Earth System Modeling simulations on petabytes of high-resolution, three-dimensional data. “This is the future of what day-to-day science research computing will look like,” he predicts.


At the Stanford Center for Genomics and Personalized Medicine (SCGPM), researchers using GCP and BigQuery can now run hundreds of genomes through a variant analysis pipeline and get query results quickly. Mike Snyder, director of SCGPM, notes, “We’re entering an era where people are working with thousands or tens of thousands or even million genome projects, and you’re never going to do that on a local cluster very easily. Cloud computing is where the field is going.”


Googlers like Fei-Fei Li, Chief Scientist for Cloud AI and ML, are excited to be able to support important research through the new avenue of the credits program: “As an academic, I’m thrilled that Google Cloud will make GCP credits available to the research community. This will help support important scientific discoveries and accelerate fundamental research that are critical for the future.”


The GCP research credits program is open to faculty doing cutting-edge research in eligible countries. We’re eager to hear how we can help accelerate your progress. If you’re interested, you can learn more on our FAQ or apply now.

Source: Google Cloud


Simplifying apps, desktops and devices with Citrix and Chrome Enterprise

As cloud adoption continues to accelerate, many organizations have found they need an ever-expanding fleet of mobile devices so that employees can work wherever and whenever they need. And research shows that when employees can work from anywhere, they can do more. According to Forbes, employee mobility leads to 30 percent better processes and 23 percent more productivity.

But as the demand for mobility grows, many organizations have also found themselves challenged by the need to provide secure mobile endpoints with access to certain legacy line-of-business or Windows apps. To help, last year we announced our partnership with Citrix to bring XenApp and XenDesktop to Chrome Enterprise.

Since bringing XenApp and XenDesktop to Chrome Enterprise, we’ve worked extensively with Citrix to help more businesses embrace the cloud. Last month, we announced that admins can now manage Chromebooks through several popular enterprise mobility management (EMM) tools, including Citrix XenMobile. And this year at HIMSS we showed how the combination of Citrix and HealthCast on Chrome Enterprise helps healthcare workers access electronic health records and virtualized apps securely on Chrome OS using their proximity badge.

All of this is the topic of an IDG webinar we’re co-sponsoring with Citrix. The webinar “Chrome OS & Citrix: Simplify endpoint management and VDI strategy” includes IDG CSO SVP/Publisher Bob Bragdon, Chrome Enterprise Group Product Manager Eve Phillips, and Citrix Chief Security Strategist Kurt Roemer as speakers, and addresses how Citrix and Chrome enable access to mission-critical business apps and create a productive workforce inside or outside corporate infrastructure.

Here’s what the webinar will cover:

  • How Chrome and Citrix can ensure secure access to critical enterprise apps.
  • How employees can be more productive through access to legacy apps in VDI. 
  • How Citrix XenApp (XA) and XenDesktop (XD) integrate with Chrome OS.
  • How Citrix’s upcoming product launches and enhancements with Chrome, GCP and G Suite can help enterprise IT teams and end users.

In March, Citrix’s Todd Terbeek shared his experiences transitioning to Chrome Enterprise, and this week Chief Security Strategist Kurt Roemer discussed how combining Citrix with Chrome can deliver expanded value across security, privacy and compliance. Our work with Citrix continues to evolve, and we’re looking forward to finding new ways to collaborate in the future.

To learn more, sign up for the webinar.

Source: Google Cloud


Meeting our match: Buying 100 percent renewable energy

A little over a year ago, we announced that we were on track to purchase enough renewable energy to match all the electricity we consumed over the next year. We just completed the accounting for Google’s 2017 energy use and it’s official—we met our goal. Google’s total purchase of energy from sources like wind and solar exceeded the amount of electricity used by our operations around the world, including offices and data centers.


What do we mean by “matching” renewable energy? Over the course of 2017, across the globe, for every kilowatt hour of electricity we consumed, we purchased a kilowatt hour of renewable energy from a wind or solar farm that was built specifically for Google. This makes us the first public Cloud, and company of our size, to have achieved this feat.


Today, we have contracts to purchase three gigawatts (3GW) of output from renewable energy projects; no corporate purchaser buys more renewable energy than we do. To date, our renewable energy contracts have led to over $3 billion in new capital investment around the world.

The road to 100 percent

We've been working toward this goal for a long time. At the outset of last year, we felt confident that 2017 was the year we'd meet it. Every year, we sign contracts for new renewable energy generation projects in markets where we have operations. From the time we sign a contract, it takes one to two years to build the wind farm or solar field before it begins producing energy. In 2016, our operational projects produced enough renewables to cover 57 percent of the energy we used from global utilities. That same year, we signed a record number of new contracts for wind and solar developments that were still under construction. Those projects began operating in 2017—and that additional output of renewable energy was enough to cover more than 100 percent of what we used during the whole year.


We say that we “matched” our energy usage because it’s not yet possible to “power” a company of our scale by 100 percent renewable energy. It’s true that for every kilowatt-hour of energy we consume, we add a matching kilowatt-hour of renewable energy to a power grid somewhere. But that renewable energy may be produced in a different place, or at a different time, from where we’re running our data centers and offices. What’s important to us is that we are adding new clean energy sources to the electrical system, and that we’re buying that renewable energy in the same amount as what we’re consuming, globally and on an annual basis.

Google's data center in Eemshaven, The Netherlands.
Google's data center in Eemshaven, The Netherlands.

Looking ahead

We’re building new data centers and offices, and as demand for Google products grows, so does our electricity load. We need to be constantly adding renewables to our portfolio to keep up. So we’ll keep signing contracts to buy more renewable energy. And in those regions where we can’t yet buy renewables, we’ll keep working on ways to help open the market. We also think every energy buyer—individuals and businesses alike—should be able to choose clean energy. We’re working with groups like the Renewable Energy Buyers Alliance and Re-Source Platform to facilitate greater access to renewably-sourced energy.


This program has always been a first step for us, but it is an important milestone in our race to a carbon-free future. We do want to get to a point where renewables and other carbon-free energy sources actually power our operations every hour of every day. It will take a combination of technology, policy and new deal structures to get there, but we're excited for the challenge. We can’t wait to get back to work.

Source: Google Cloud


Expanding our cloud network for a faster, more reliable experience between Australia and Southeast Asia

Earlier this year, we announced that we are expanding our global infrastructure with new regions and subsea cables, advancing our ability to connect the world and serve our Cloud customers with the world’s largest network.

Today, we’re excited to announce our investment in the Japan-Guam-Australia (JGA) Cable System.  

This new addition to the Google submarine network family, combined with investments in the  Indigo, HK-G and SJC subsea cables,will give GCP users access to scalable, diverse capacity on the lowest latency routes via a constellation of cables forming a ring between the key markets of Hong Kong, Australia and Singapore. Our investment in these cables builds on our other APAC cable systems, namely Unity, Faster and PLCN, interconnecting the United States with Japan, Taiwan and Hong Kong.

Taken together, these cable investments provide improved connectivity to GCP’s five cloud regions across Asia and Australia (with more on the way),so that companies using GCP can serve their customers no matter where they are.

The JGA cable system will have two fiber pairs connecting Japan to Guam, and two fiber pairs connecting Guam to Sydney. This provides deeply scalable capacity to both our users and Google Cloud Platform customers. JGA is being co-built by NEC Corporation and Alcatel Submarine Networks. The JGA-South segment is being developed by a consortium of AARnet, Google, and RTI-C, while the JGA-North segment is a private cable being developed by RTI-C. Together, the segments will stretch 9500 km (or nearly 6000 miles).

Whether we’re delivering directions to Maps users, videos to YouTube viewers, or GCP services to businesses, we know a fast and reliable infrastructure makes all the difference. That’s why we continue to invest in strategic routes, many of which require crossing oceans. To learn more about our network, visit our website.

JGA-1

Source: Google Cloud


ICYMI in March: here’s what happened in G Suite

Just like that, another month down.

In March, we announced a slew of security updates to Google Cloud, including enhancements to G Suite. In a nutshell, G Suite companies can now use advanced configurations to help fend off phishing scams. These updates will continue to help businesses block (ph)ishy activity, like if an untrusted sender tries to share encrypted attachments or if someone tries to trick you by sending information from a domain that looks like yours.

With these protections in place, more than 99.9 percent of Business Email Compromise (BEC) scenarios—when someone impersonates an executive to get sensitive information—are automatically moved to spam or flagged to users as shifty. Sorry, Charlie.

We also automatically enabled basic device management for mobile devices that access G Suite. Now IT admins can better enforce pass codes, erase confidential data for Android and iOS devices with selective account wipe and more without users needing to install profiles. Lastly, we added IRM controls to Team Drives to prevent folks from printing, downloading or copying files they shouldn’t have access to.


Many of these protections are default-on, which means you don’t have to do a thing. Read up here, or get started using the security center for G Suite.
ICYMI 1
ICYMI 2

This one is so simple. Did you know that you can make a copy of a Google Doc or Sheet with a quick URL change? In the URL of your document, delete the information before the final backslash. In this case, change “edit” to “copy.” Done!

And since security should always be top of mind, brush up on how to manage your share settings in Docs or other apps on our Help Center.

ICYMI 2
People predictions

→Most of us track down files in Google Drive by searching for the name of the person who shared a file with us. Because of this, Drive is going to start intelligently organizing the “Shared with Me” section by listing names and the files that people have shared with you, so you can track down files faster.

→ Two-step verification is an easy and effective way to protect G Suite users, which is why we recommend that businesses use security keys. Moving forward, all G Suite admins—not just G Suite Business admins—will be able to manage the deployment of security keys and view usage reports. Learn more.

→ Now your jams in Jamboard will automatically save to Drive.

Source: Google Cloud


Noodle on this: Machine learning that can identify ramen by shop

There are casual ramen fans and then there are ramen lovers. There are people who are all tonkatsu all the time, and others who swear by tsukemen. And then there’s machine learning, which—based on a recent case study out of Japan—might be the biggest ramen aficionado of them all.


Recently, data scientist Kenji Doi used machine learning models and AutoML Vision to classify bowls of ramen and identify the exact shop each bowl is made at, out of 41 ramen shops, with 95 percent accuracy. Sounds crazy (also delicious), especially when you see what these bowls look like:
Ramen bowls made at three different Ramen Jiro shops.
Ramen bowls made at three different Ramen Jiro shops

With 41 locations around Tokyo, Ramen Jiro is one of the most popular restaurant franchises in Japan, because of its generous portions of toppings, noodles and soup served at low prices. They serve the same basic menu at each shop, and as you can see above, it's almost impossible for a human (especially if you're new to Ramen Jiro) to tell what shop each bowl is made at.


But Kenji thought deep learning could discern the minute details that make one shop’s bowl of ramen different from the next. He had already built a machine learning model to classify ramen, but wanted to see if AutoML Vision could do it more efficiently.


AutoML Vision creates customized ML models automatically—to identify animals in the wild, or recognize types of products to improve an online store, or in this case classify ramen. You don’t have to be a data scientist to know how to use it—all you need to do is upload well-labeled images and then click a button. In Kenji’s case, he compiled a set of 48,000 photos of bowls of soup from Ramen Jiro locations, along with labels for each shop, and uploaded them to AutoML Vision. The model took about 24 hours to train, all automatically (although a less accurate, “basic” mode had a model ready in just 18 minutes). The results were impressive: Kenji’s model got 94.5 percentaccuracy on predicting the shop just from the photos.

Confusion matrix of Ramen Jiro shop classifier by AutoML Vision

Confusion matrix of Ramen Jiro shop classifier by AutoML Vision (Advanced mode). Row = actual shop, column = predicted shop. You can see AutoML Vision incorrectly identified the restaurant location in only a couple of instances for each test case.

AutoML Vision is designed for people without ML expertise, but it also speeds things up dramatically for experts. Building a model for ramen classification from scratch would be a time-consuming process requiring multiple steps—labeling, hyperparameter tuning, multiple attempts with different neural net architectures, and even failed training runs—and experience as a data scientist. As Kenji puts it, “With AutoML Vision, a data scientist wouldn’t need to spend a long time training and tuning a model to achieve the best results. This means businesses could scale their AI work even with a limited number of data scientists." We wrote about another recent example of AutoML Vision at work in this Big Data blog post, which also has more technical details on Kenji’s model.


As for how AutoML detects the differences in ramen, it’s certainly not from the taste. Kenji’s first hypothesis was that the model was looking at the color or shape of the bowl or table—but that seems unlikely, since the model was highly accurate even when each shop used the same bowl and table design. Kenji’s new theory is that the model is accurate enough to distinguish very subtle differences between cuts of the meat, or the way toppings are served. He plans on continuing to experiment with AutoML to see if his theories are true. Sounds like a project that might involve more than a few bowls of ramen. Slurp on.

Source: Google Cloud