Cloud

As we enter the holiday season, retailers are working behind the scenes to ensure they can provide the best experiences for customers, in store and online. Challenges in retail do not begin or end during the holiday season as sudden shifts in customer preferences, supply chain nuances, and overall demand ebbs and flows take place year round and retailers must be prepared to adapt swiftly. 

Google Cloud’s retail customers globally, in total, saw more online traffic in the first six months of 2022 than all of 2019. This year, retailers can expect an early launch to holiday shopping activities, as 50% of consumers plan to start purchasing goods before the traditional Black Friday kick-off.  

The very same improvements made to automate and improve retail infrastructure can prepare it for holiday surges and support year-round innovation. Let’s take a look at how Ulta Beauty, the largest beauty retailer in the U.S., is partnering with Google Cloud, MongoDB Atlas, commercetools, and HCLTech to cover these two areas and more. 

Architecting for innovation

Creating personalized shopping experiences in stores and online is key to Ulta Beauty’s success. This commitment is best demonstrated through Ulta Beauty’s Virtual Beauty Advisor. Built on Google Cloud, this tool enhances shoppers’ experiences with personalized recommendations in addition to the ability to try on makeup virtually with GLAMLab.

As innovators in support of the best possible guest experience, Ulta Beauty needed to re-architect its infrastructure for greater agility and stability. 

To start, Ulta Beauty chose to use Google Kubernetes Engine (GKE) as the backbone and orchestrator to build and deploy cloud-native applications. The Google Cloud deployments coincided with an organizational move from end-to-end application development to one that focuses on individual features, specific modules, and micro-applications.

This strategic change allowed Ulta Beauty to fix bugs, experiment with new offerings, and drive customer experiences faster and more efficiently. Thanks to the transformation and GKE, Ulta Beauty’s developer team now accelerates time to market for new products and services, and delivers new ways to engage with customers more quickly. These efforts all ladder to create ‘WOW’ experiences for the retailers’ guests who have emotional and personal connections to beauty and wellness. They can now discover and experience products that are served to them based on individual preferences. 

Adapting to the new environment comes with its own set of challenges. “Microservices are not a silver bullet,” says Sethu Madhav Vure, IT Architect, Ulta Beauty. “For Ulta Beauty, the biggest challenge was how to break up a monolithic environment into multiple applications. We had to evolve our core systems—without impacting today’s services—and address what was needed for the future.”

Google Cloud partner HCLTech provided expert guidance throughout the re-architecting process, defining the solution blueprint and cloud-native deployment architecture through cross-functional workshops. HCLTech then assisted with the actual migration and platform setup, paving the way for fully automated, continuous integration and continuous delivery (CI/CD) pipelines to support faster rollouts and deployment architecture to drive higher availability and scalability.

Ulta Beauty took a domain-driven design approach to identify operations that could be grouped together to reduce complexity and improve scalability. Now, the applications are based on multiple domains, such as Commerce, Promotions, Catalog, Order, Customer, and Inventory. The new architecture prompted a fresh look at storage requirements to scale dynamically alongside its modernized applications.

For Ulta Beauty, MongoDB Atlas proved to be the best database solution for dynamic scaling, ease-of-use, and integrations with Google Cloud. The company also leveraged an entry-level plan to prove the value of MongoDB Atlas before investing in the technology. 

“MongoDB Atlas offers a free tier that gave us an opportunity to quickly demonstrate tangible benefits of a proof of concept,” says Vure. “Once we proved the value of MongoDB Atlas, we benefited from the straightforward resource allocation supported by Google Cloud and MongoDB.”

Integrations between MongoDB Atlas and Google Cloud allow Ulta Beauty to take an iterative approach to new projects. The company creates new clusters in an existing project, then piggybacks them onto an existing Private Service Connect setup between a MongoDB project and Google Cloud project.

By removing complexities within infrastructure management, Ulta Beauty can manage its incredible amount of data, such as member preferences and purchases,  that fuels its event-driven architecture. The much more agile infrastructure enables Ulta Beauty to deploy and scale offerings faster than ever.

“We recently had an unplanned traffic surge that impacted our domain services. It took less than an hour for MongoDB Atlas to scale up to the next level of the cluster and manage that traffic,” says Vure. “The on-demand, dynamic scaling, plus GKE, has saved the day more than once.”

Preparing for a happy holiday season

This holiday season, Ulta Beauty has a stronger technical foundation to manage demand surges and provide customers seamless shopping experiences. Previously, the company used 50 pods in a cluster, each with 6 GB of RAM without domain stores, to handle about 100 transactions each second. With domain stores, the same 6 GB of RAM with just 20 GKE pods was able to scale up to 2,400 transactions per second. 

With Google Cloud as its technology foundation, Ulta Beauty partnered with Google Cloud partner commercetools to evolve its application APIs as products and properly separate interfaces and capabilities.

Ulta Beauty uses event-based integrations within commercetools to identify how best to leverage Cloud Pub/Sub middleware on top of MongoDB Atlas integrations. Patterns established here were extended into MongoDB change streams and in turn improved business processes.

“Working with the right technology partners has helped us to avoid analysis paralysis that can happen when developer teams spend a lot of time trying to understand and manage every detail,” says Vure. “Instead, we convert a proof of concept into a working solution, and quickly bring it to market. It’s been a major shift in our IT culture as we try out new things weekly and see incredible support from leadership.”

The improvements enable Ulta Beauty to maintain a high level of innovation, performance, and customer service year-round. Now, when the holiday shopping season begins, Ulta Beauty is prepared to handle surges in traffic through auto-scaling with Google Cloud and MongoDB Atlas. Customers get what they want, when they want, free from the frustrations of outages. 

“With these changes, we are ready for a holiday season that everyone–even those of us in IT—gets to enjoy. We’re positioned to continuously focus on new, better ways to serve our guests,” says Vure.

Check out MongoDB and commercetools on Google Cloud Marketplace to learn more about what these partners can do for your business. 

The content of this blog was taken from a Google Cloud interview with Ulta Beauty.

These days, many stores can’t keep their shelves stocked with the basic items people need at an affordable price. Although the retail industry has always struggled to manage inventory efficiently, the COVID-19 pandemic has created supply chain disruptions and staffing shortages that make it even more difficult for stores to meet customer expectations.

That’s why we builtZippedi. With the Zippedi AI inventory robot and mobile app, any store can scan its shelves and efficiently identify, replace, and reorganize products that are running low or out of stock. Zippedi automatically creates and delivers detailed inventory reports to store managers—and alerts employees when items, price tags, and signs need to be restocked or updated.

Building an automated inventory management platform

Zippedi began its journey in 2018 as an engineering project at the Pontificia Universidad Católica de Chile. Although we were determined to create an automated inventory management platform, we knew we couldn’t effectively scale or accelerate time to market without a reliable technology partner. Zippedi is a company that develops high-tech software and hardware solutions. Scaling a startup is not trivial with too many flanks to address. On-premises or self-hosted/managed resources for software development & deployment would’ve meant to create further dependencies on hardware, not only for clients (by building robots) but also datacenters to process all the incoming data robots are gathering worldwide. 

We looked at the options and identifiedGoogle Cloud, including theGoogle for Startups Cloud Program, as the best choice to help us reduce IT management costs, affordably expand our technology stack, and allocate more resources to R&D. In just a year, we migrated on-premises IT infrastructure to Google Cloud, built a fully functioning AI robot, coded an easy-to-use mobile app, and tripled the size of our team.

We now leverage thesecure-by-design infrastructure of Google Cloud to rapidly process and analyze millions of shelf scans for stores around the world with an incredibly high level of precision. Specifically, we rely onVertex AI to quickly build, deploy, and scale the advanced machine learning (ML) models that classify products, identify items running low or out of stock, and spot outdated price tags and signs.

We useCompute Engine to run Zippedi’s online prediction and inference engine on scalable virtual machines that lower computing costs. These cost savings enable us to shift more of our budget to R&D and rapidly develop new solutions and services for our customers. We’re also seeing significant cost savings withBigQuery, which gives us a 26% to 34% lower three-year TCO than cloud data warehouse alternatives.

We use additional Google Cloud solutions to power Zippedi. We automate database provisioning and storage capacity management withCloud SQL, run large analytical and operational workloads onCloud Bigtable, develop and deploy highly scalable containerized applications onCloud Run, and author, schedule, and monitor pipelines across multiple clouds withCloud Composer. Lastly, we develop and introduce new features for our Zippedi mobile app onApp Engine.

Scaling Zippedi with the Google for Startups Cloud Program

Building, scaling, and deploying a market-ready automated inventory management platform in just a year would not have been possible without the incredible support of the Google for Startups Cloud Program.

For example, the dedicated Startups Success Team made sure that our customized microservices and ML models seamlessly migrated to Google Cloud. We also used Google Cloud credits to affordably trial differentAI and machine learning products and Google Cloud databases. We continue to tap the deep technical knowledge of the Google Cloud team and technical experts to further optimize our automated inventory management platform with additional Google Cloud solutions. These could potentially include MLOps pipelines through Vertex AI,Cloud TPUs, andLooker. 

Since 2020, the retail industry has struggled to keep store shelves stocked with the basic items people need at an affordable price. With Zippedi, stores are successfully mitigating supply chain disruptions and lowering costs by efficiently identifying, replacing, and reorganizing products that are running low or out of stock.

Today, many stores around the world rely on Zippedi to stabilize consumer prices by streamlining inventory management, increasing employee efficiency, and lowering operating costs. Our customers are seeing impressive results as they eliminate pricing errors by 90%, improve on-shelf availability by over 10%, and bolster in-store sales by up to 5%.

As we expand our team, we look forward to launching Zippedi in new markets and empowering retailers to overcome the delivery delays and staffing shortages caused by the COVID-19 pandemic. We will introduce new features and services to help the retail industry mitigate supply chain disruptions and make it easier and more affordable for people to buy the basic items they need to thrive.

If you want to learn more about how Google Cloud can help your startup, visit our pagehere to get more information about our program, and sign up for our communications to get a look at our community activities, digital events, special offers, and more.

Tech companies are ramping up the search for highly skilled data analytics, AI and ML professionals, with the race to AI accelerating the crunch.1 They are looking for cloud experts  who can successfully build, test, run, and manage complex tools and infrastructure, in roles such as data analysts, data engineers, data scientists, and ML engineers. This workforce takes vast amounts of data and  puts it to work solving top business challenges, including customer satisfaction, production quality and operational efficiency. 

Learn about the business impact of data analytics, ML and AI skills

Find out how Google Cloud ML, AI and data analytics training and certification can empower your team to positively impact operations in our latest IDC Business Value Paper, sponsored by Google. Key findings include:

69% improvement in staff competency levels

31% greater data accuracy in products developed

29% greater overall employee productivity

Download the latest IDC Business Value Paper, sponsored by Google, “The Business Value of Google Cloud ML, AI, Data Analytics Training and Certification.” (#US48988122, July 2022).

Google Cloud customers prioritize ML, AI and data training to meet strategic organizational needs

Our customers are seeing the importance and impact of data analytics, AI and ML training on their teams and business operations. 

The Home Depot (THD) upskilled staff on BigQuery to derive business insights and meet customer demand, with 92% reporting that training was valuable, and 75% confirming that they used knowledge from their Google Cloud training on a weekly basis.2

THD was challenged with upskilling IT staff to extract data from the cloud in support of efficient business operations. Additionally, they were working on a very short timeline (weeks as opposed to years) to train staff to enable a multi-year cloud migration completion. This included thousands of employees and a diverse range of topics. Find out how they successfully executed this major undertaking by developing a strategic approach to their training program in this blog.

LG CNS wanted to grow cloud skills internally to provide a high level of innovation and technical expertise for their customers. They enjoyed the flexibility and ability to tailor content to meet their objectives, and have another cohort planned.3

Looking to drive digital transformation and solution delivery, LG CNS partnered with Google Cloud to develop a program that included six weeks of ML training through the Advanced Solutions Lab (ASL). Read the blog to learn more about their experience.

Gain the latest data analytics, ML and AI skills on Google Cloud Skills Boost

Discover the latest Google Cloud training in data analytics, ML and AI on Google Cloud Skills Boost. Explore the role based learning paths available today which include hands-on labs and courses. Take a look at the Data Engineer, ML Engineer, Database Engineer and Data Analystlearning paths today for you and your team to get started on your upskilling journey. 

To learn about the impact ML, AI and data analytics training can have on your business, take a look at the IDC Business Value Paper, available for download now.

1. Tech looks to analytics skills to bolster its workforce
2. THD executed a robust survey directly with associates to gauge the business gains of the training program. Over the course of two years, more than 300 associates completed the training delivered by ROI Training.
3. Google Cloud Learning services’ early involvement in the organizational stages of this training process, and agile response to LG CNS’s requirements, ensured LG CNS could add the extra week of MLOps training to their program as soon as they began the initial ASL ML course.

Saving money never goes out of style. Today, many of our customers use Compute Engine resource-based committed use discounts (CUDs) to help them save on steady-state compute usage within a specific machine family and region. As part of our commitment to offer more flexible and easy ways for you to manage your spend, we now offer a new type of committed use discount for Compute Engine: flexible CUDs.

Flexible CUDs are spend-based commitments that offer predictable and simple flat-rate discounts (28% off 1-year, and 46% off 3-years) that apply across multiple VM families and regions. Similar to resource-based CUDs, you can apply flexible CUDs across projects within the same billing account, and to VMs of different sizes and tenancy, to support changing workload requirements while keeping costs down. 

Today, Compute Engine flexible CUDs are available for most general-purpose (N1, N2, E2, N2D) and compute-optimized (C2, C2D) VM usage, including instance CPU and memory usage across all regions (refer to the complete list with more VM families to come). 

Similar to resource-based CUDs, flexible CUDs are discounts over usage, not capacity reservations. To ensure capacity availability, make a separate reservation, and CUDs will apply automatically to any eligible usage as a result.

You can purchase CUDs from any billing account, and the discount can apply to any eligible usage in projects paid for by that billing account. When you purchase a flexible CUD, you pay the same commitment fee for the entirety of the commitment term, even if your usage falls below this commitment value. The commitment fee is billed monthly. Once a commitment is purchased, it cannot be canceled.

For the best combination of savings and flexibility, you can combine resource-based CUDs and flexible CUDs together. You can have standard resource-based CUDs to cover your most stable resource usage and flexible spend based CUDs to cover your more dynamic resource usage. Every hour, standard CUDs apply first to any eligible usage followed by flexible CUDs, optimizing the use of your CUDs. Finally, any usage overage or usage that’s not eligible for either type of CUDs, will be charged based on your on-demand rates. 

Here is a quick summary of the differences between resource based CUDs and flexible CUDs

“As Lucidworks expands our product offerings, Google Cloud’s Flexible CUDs have been the perfect solution to optimize cost while giving us the flexibility to shift workloads to different regions based on customer demographics and different instance families based on performance characteristics.” — Matt Roca, Director of Cloud Governance and Security, Lucidworks

Understanding flexible CUDs 

You can purchase a flexible CUD in the Google Cloud console or via the API. A flexible CUD goes into effect one hour after purchase, and the discounts will automatically be applied to any eligible usage. 

Your flexible CUD is applied to eligible on-demand spend by the hour. If during a given hour, you spend less than what you committed to, you will not fully utilize your commitment or realize your full discount.

For example: If you want to cover $100 worth of on-demand spend every hour by a flexible CUD, you will pay $54/hour (46% off) for 3 years (payable monthly), and receive a $100 credit that applies automatically to any eligible on-demand spend. The $100 credit burns down at the eligible on-demand rate for every eligible SKU, and expires if unused.

Attributing flexible CUDs credits
If you are running multiple projects within the same billing account, the credits from flexible CUDs are attributed proportionally across projects within the billing account and across SKUs within the same project according to their usage proportion.

Planning for flexible CUDs purchases
A good way to think about how to purchase and use resource based CUDs with flexible CUDs is to first forecast and purchase resource based CUDs based on your steady state resource spend, to get the deepest discounts. A best practice is to use flexible CUDs for more variable and growing workloads, and to use on-demand VMs, or Spot VMs, for the rest of your usage. 

Get started with flexible CUDs today

Building a business in the cloud can be complicated; paying for it should be easy. We designed flexible CUDs to make it easy for organizations to enjoy significant discounts across a wide variety of Google Cloud resources in a way that’s simple and predictable. For more details on how to purchase and use flexible CUDs and to get started, refer to this documentation.

Security insights are a critical piece of a secure enterprise browsing solution.Integrating your chosen security solution with Chrome, made possible by Chrome Enterprise Connectors Framework, gives IT and security teams a birds-eye view of the potential threats users face on the web and make data-driven decisions in their security journey.

We’re pleased to share that the Chrome and CrowdStrike Falcon LogScale integration is now available in the Google Admin console in Chrome Browser Cloud Management. Organizations can now easily set up the integration to extract and analyze insights from security events within managed Chrome browsers. 

Available Security Events

Malware transfer

Password changed

Unapproved password reuse

Unsafe site visit

Log-in events

Password breaches 

Extension installs (coming soon)

Crash events (coming soon)

Enrolling machines in Chrome Browser Cloud Management

Getting started is easy. The first step is to make sure Chrome Browser Cloud Management is set up for your organization. This tool helps manage Chrome browser from a single, cloud-based Admin console across Windows, Mac, Linux, Android and iOS at no additional cost. This is the same console where IT teams can also manage Chrome OS. 

Check out this guide for steps on how to enroll your devices. Once you are done, or if you already have Chrome Browser Cloud Management in place, move to the steps below. 

Setup in CrowdStrike Falcon LogScale instance

Log into your CrowdStrike Falcon LogScale instance. 

You will need to create a new repository for your Google Chrome data. 

Once you’ve created a new repository, under packages, click on the Marketplace section. 

Search for Chrome, then install the CrowdStrike Falcon LogScale package for google/chrome-enterprise.When choosing the package, the README provides information about the package contents and other related information.

After installing the package, under Ingest, click on Ingest tokens and click the Add token button. Give it a name and type in Chrome under the assigned parsers. 

Select the Google_Chrome_Enterprise parser. Click on the eye icon to bring up the token value.

Hit the copy button and save the token value, as you will be entering this into the admin console in the following section.

Log into the Google Admin console at admin.google.com and select the organizational unit containing the enrolled browsers you want to send security events to CrowdStrike. 

Navigate to Devices>Chrome>Users and browsers. Add a filter for “event reporting.”

Under events reporting, select Enable event reporting. Under the additional settings, you can specify which events you want to send to CrowdStrike.

Now that the events are turned on, click on the blue hyperlink called “Reporting connector provider configurations” to take you to the connector provider configurations, or it can be found under Devices>Chrome>Connectors.

Click the New Provider Configuration button and select CrowdStrike as the provider.

Enter the configuration name that you want this connector to display as in the Google Admin console.

Enter the host name of your CrowdStrike instance (most likely cloud.us.humio.com or sa-cluster.humio-support.com or your on-prem instance) and the ingest token value from step 7 of the last section

Press the Add Configuration to save.

Select the Organizational Unit where the  reporting events are turned on, select the Chrome CrowdStrike connector created in the previous step, and hit Save.

You can also download the setup guide here or watch the setup of integration with CrowdStrike Falcon LogScale end-to-end in the video below:

Chrome is continuously working with other partners to expand the security ecosystem to help businesses work safer on the web. Learn more about Chrome browser securityhere.