Jet.com - E-commerce challenger eyes the top spot, runs on the Microsoft cloud
Marc Lore is perhaps best known as the creator of the popular e-commerce site Diapers.com, which was eventually sold to Amazon. Now, the entrepreneur and his team are ready to compete head-on with the e-retailing giant through an innovative online marketplace called Jet.com. To get up and running quickly, Jet built its entire e-commerce platform, including development and delivery infrastructure, on Microsoft Azure, using both .NET and open-source technologies.
Business Challenge
In 2010, Marc Lore sold his company Quidsi (which ran e-retailing sites like Diapers.com and Soap.com) to Amazon for $550 million. Four years later, Marc is competing against Amazon directly—with the creation of a new online marketplace called Jet.com.
There are many reasons to think that Lore might just pull it off. For one, he plans to eliminate any margins from product sales. The company’s only source of revenue will come from membership dues, eliminating the kind of mark-ups that Amazon charges and passing the savings on to the customer. In addition, an innovative pricing engine will work to reduce or eliminate costs in the e-commerce value chain, especially fulfillment costs and marketplace commissions.
“Our pricing engine will continually work out the most cost-effective way to fulfill an order from merchant locations closest to the consumer,’ explains Lore, Co-Founder and CEO of Jet. “The engine will also figure out which merchants can fulfill most cheaply by putting multiple requested items into one shipment. And so we can cut probably 10 percent of a cost of a typical e-commerce transaction just by being smarter about fulfillment.”
With a value proposition like that, the company is confidently looking forward to explosive growth. “We want to be one of the leading e-commerce destinations in a very short period of time—18 to 36 months,” says Mike Hanrahan, CTO at Jet.
To fuel that growth, Jet was able to quickly secure more than $220 million in financing. Meanwhile, aggressive marketing has already created a user base of more than 400,000 customers—even before the site launched.
Next step: find the right cloud partner to support the company’s ambitious growth plan. “We realized that we simply did not have the resources to build and manage the kind of datacenters and development infrastructure to meet our growth strategy,” says Hanrahan. “So we quickly decided on a cloud model.”
Solution
The decision to work with Microsoft Azure was driven, in part, by the .NET development platform—and Visual F# in particular—which was a good fit with the microservices architecture used to build Jet. As Hanrahan explains, “The event-driven, microservices paradigm eliminated a lot of the overhead that comes with a service-oriented architecture such as Amazon uses, meaning that everyone can build all their systems in parallel and then publish and subscribe to an event bus. We found that F# works very well with this paradigm, especially the immutable data streams that are a key part of our microservices architecture.”
The Microsoft Visual Studio development system is the primary IDE for back-end infrastructure, with Node being used on the front end. To get its code through development and into production as fast as possible, Jet uses a mix of Azure App Service, Azure Web Roles and custom servers, with deployment happening from Jenkins.
Jet also has many open-source middleware components, which it runs on Azure Virtual Machines, including Elasticsearch, RedisLabs, Hadoop, and Event Store—an open-source event-sourcing data store.
Jet is taking advantage of several other Azure services to streamline its development processes. For example, to make it easier for merchant partners to integrate with the platform, it has created a developer portal for its APIs using Azure API Management.
Jet is also using Azure Key Vault to store encryption keys as well as Azure Application Insights, which will provide real-time alerts to its developers to help them identify and triage problems as they occur. Application Insights also enables Jet to learn, in real time, how customers are using their application, so they can implement an agile build-measure-learn cycle.
“Being able to leverage so many off-the-shelf services and tools from Azure enabled us to go from zero to a full- fledged e-commerce marketplace in just about 12 months. That same system would have taken us at least two years to build on our own, plus capex costs,” says Hanrahan.
The company also relies on other Microsoft cloud services to run its day-to-day business, including Office 365 and Azure Active Directory. In fact, Jet’s entire operation is now run in the cloud using Azure. “We have no servers at all onsite right now, not a single one,” says Hanrahan.
Benefits
Working with Microsoft Azure cloud services has provided Jet with a level of flexibility and scalability that has been critical to its aggressive development schedule.
Moving from code to production in minutes.
By using App Service for its consumer front end, Jet has been able to dramatically streamline its development process, so that it can build, deploy, and scale consumer-grade web apps more rapidly. As Hanrahan says, “We’ve been able to get our critical code through our CI/CD process in a couple of minutes using App Service.”
Scaling automatically to meet customer demand
As with any popular e-retailing site, Jet requires extremely rapid and flexible scaling based on ever-changing customer traffic. To streamline this process, Jet was able to set up auto-scaling on both PaaS servers and App Service to scale its servers based on load or schedule. “Because both PaaS and App Service scale automatically for us, we are able to throw as many machines as we need at the front end, when we need them,” says Hanrahan.
Accommodating rapidly growing storage requirements
As the company grows, Azure provides a wide range of storage options to handle virtually any amount of data. “Right now our data warehouse sits in a SQL Server instance, but we will be augmenting that using HD Insight,” says Hanrahan. Azure HDInsight is designed to handle any amount of data, scaling from terabytes to petabytes on demand.
With Azure, Jet has created a cloud infrastructure that’s ready to meet the company’s most ambitious growth plans. “To be one of the best e-commerce destinations in the US, we will have to handle millions of customers, placing tens of thousands of orders a day. That requires a top-class e-commerce system built on a flexible, open cloud platform. That is exactly what we got with Azure,” says Hanrahan.
Source: https://customers.microsoft.com
Read customer success stories to learn how Managed Solution helps businesses implement technology productivity solutions.
New Azure SQL Database offerings add capabilities to scale data up and out
As written by Tiffany Wissner, Senior Director of Data Platform Marketing on Technet.com.
At Microsoft we’re committed to meeting the needs of our customers by offering choices as they build cloud applications for scale and performance. Today we’re continuing to deliver on this commitment with two announcements:
•New high performance levels for Azure SQL Database, P4 and P11, with greater flexibility to start small and scale up as data grows
•Basic and Premium tier for Elastic Databases for more flexibility for SaaS developers looking to scale out with higher performance and more resource control.
As more customers build more powerful applications in Azure using Azure SQL Database, we've seen strong demand for more options to scale performance. We’re dedicated to offering a portfolio of data services for cloud applications that effectively scales up to support data growth from more users, and scales out to support growth from more customers and fluctuating resource demands.
To meet this scale up demand, we are announcing the release of new Azure SQL Database Premium performance levels, P4 and P11. With the introduction of both P4 and P11, we’ve expanded the range of throughput performance developers can leverage to scale their applications. Developers can quickly start with our basic or standard performance tiers and easily scale database performance up to 350 times greater with our premium tier as their data needs grow – all without code or application changes. This enables developers to match application needs with the right database performance.
On the scale out front, we have seen an increasing demand for Elastic Database pools since we announced its public preview. We’ve seen SaaS developers adopt and embrace elastic databases because it makes it easier to manage their growing number of customers. With Elastic Database pools SaaS developers can manage hundreds of databases for each of their customers, optimize the performance of each database, and manage the cost of these databases to support their explosive growth. This means their customers still get a great experience, even during peak usage periods, and there’s no need to over-provision resources during lulls.
One example is StudioPlus, which offers solutions for professional photographers. CEO Matthew Hunt says, “We are expanding our legacy on-premises solution with a SaaS solution. Azure SQL Database is a fundamental part of this journey because it allows us to leverage an existing code base and gives us a platform for unlimited customer growth. From a DevOps perspective, it’s simple to set up and manage Azure SQL Database Elastic Database pools to support the unlimited scale-out of a one-database-per-customer pattern.”
To support the increasing demand of Elastic Database Pools, we are now introducing a public preview of two additional service tiers for Elastic Databases, Basic and Premium. These new tiers give organizations and developers with SaaS applications greater options and flexibility for getting started with Elastic Databases and scaling out to meet new customer demands.
We are thrilled that we can support StudioPlus and a growing number of developers on their SaaS journey with Azure SQL Database.
Our announcements today add scale and performance options, and these releases are just part of our end-to-end data platform. We continue to make it easier for customers to maximize their data dividends with our data platform and services. It’s never been easier to capture, transform, mash-up, analyze and visualize any data, of any size, at any scale, in its native format using familiar tools, languages and frameworks in a trusted environment on-premises and in the cloud.
