SHOPLINE Buys Time and New Customers with Serverless on AWS
2021
By building its platform on AWS, SHOPLINE adopts modern technology such as containers to quickly scale its business and infrastructure while minimizing cost. SHOPLINE is an omnichannel smart commerce platform based in Hong Kong that helps global customers bring their businesses online. It uses AWS Lambda as a serverless compute engine, Amazon EKS to run Kubernetes applications, and AWS Shield Advanced to protect against DDoS attacks.
Managed services from AWS allow us to work at full speed without worrying about cost, workforce, and
security, or constantly acquiring knowledge about emerging technologies.”
Ronald Li
Engineering Director, SHOPLINE
Empowering 250,000 Merchants to Sell Online
For small businesses without a brick-and-mortar presence, setting up an e-commerce site can help boost sales
and reach customers in new markets. However, building and maintaining a successful site is easier said than
done. The global
average buyer conversion
rate is just three percent, so it’s important to develop a highly secure and available e-commerce site
that attracts and retains customers.
SHOPLINE is a global smart commerce enabler with a mission to empower
merchants to succeed everywhere. Since its founding in 2013, the Hong Kong-based startup has helped more
than 250,000 merchants (data up until 2021) across Southeast Asia and the United States set up their online
stores serving over 520 million customers. In 2019, SHOPLINE acquired point of sale (POS) systems that it
modified to support offline-to-online (O2O) business strategies and expand its customer base to enterprise
businesses with physical locations.
Working at Full Speed, Following Best Practices
SHOPLINE launched its platform on Amazon Web Services (AWS), selecting the cloud provider for various
reasons. “Managed services from AWS allow us to work at full speed without worrying about cost, workforce,
and security, or constantly acquiring knowledge about emerging technologies,” says Ronald Li, engineering
director at SHOPLINE. “We can consult with and receive technical support from the AWS team, and benefit from
sharing use cases from other businesses to avoid common pitfalls when implementing new technology.”
SHOPLINE uses
Amazon Elasticsearch Service as the primary
search engine on the backend and front end of its customers’ e-commerce sites. It relies on
Amazon ElastiCache to deliver fast response times to
searches, especially for websites supporting visitors from different countries. Using
Amazon CloudFront as a secure content delivery network is
also key in enabling low latency across borders. SHOPLINE leverages
AWS Shield Advanced to detect and mitigate sophisticated
cyberattacks such as distributed denial-of-service (DDoS) attempts.
Accelerating Innovation with Serverless Infrastructure
The SHOPLINE platform began with a traditional infrastructure on the AWS Cloud, with engineers deploying
manually via virtual servers. It has since grown into a fully-automated DevOps operation with a continuous
integration/continuous deployment (CICD) pipeline. “One of the things that has kept us on AWS is the ongoing
rollout of new services integrated into our existing AWS environment,” Ronald says. “We have frequent
technical workshops to update us on new services as they’re launched, which allows us to keep the innovation
wheel moving.”
In one such workshop, SHOPLINE engineers realized the benefits of serverless technology and containers. The
company now uses
AWS Lambda to autoscale containerized applications on
Amazon
Elastic Kubernetes Service (Amazon EKS). AWS Lambda also plays an integral role in automating
permissions and access control. This allows SHOPLINE’s DevOps engineers to spend 80–90 percent of their time
on development.
Ronald estimates that without such tools and managed services on the AWS Cloud, engineers would have to
dedicate 50 percent of their time to infrastructure management and administrative tasks, leaving
considerably less time for development. Furthermore, if the team had tried to build their own
production-grade Kubernetes clusters with supporting DevOps tools, it would’ve taken a month to launch the
first cluster. With support from AWS and Amazon EKS as a managed service, deployment takes just 15 minutes.
Focusing on Differentiated Development Work
SHOPLINE has experienced rapid growth since launching, particularly in the two years since acquiring its POS
system in 2019. During this period, the company more than doubled the number of merchants supported. As is
the case for many successful startups, business expansion has outpaced hiring.
Allowing engineers to focus on what they enjoy—creative development work versus tedious maintenance or
approval processes—has allowed SHOPLINE to retain and recruit skilled team members. While the business has
an experienced IT team, it has leveraged
AWS Infrastructure Event Management (IEM)
to support scaling in the run up to major shopping events and fine-tuning its IT infrastructure, given
budget and time constraints.
Leveraging Containers for Quick Scaling at Low Cost
Scalability is a critical part of SHOPLINE’s platform and its value proposition to customers. Implementing
containers and Kubernetes orchestration enables fast scaling during major sales events such as Black Friday
and throughout time-bound sales promotions. During such sales, customers receive sizable discounts within a
short time window, which often results in e-commerce traffic spiking up to 3 times.
Scaling to support peaks on traditional virtual servers can take 3–4 minutes, Ronald relates, but with
containerized deployment it takes less than 20 seconds. In 2020, for example, when the first wave of the
pandemic hit, SHOPLINE supported a 50 percent increase in its merchants’ traffic volumes with containerized
architecture, scaling up in near real-time. To further reduce the time to scale, SHOPLINE recently
introduced
Amazon Fargate as a serverless compute engine for containers.
“We’re confident that Amazon Fargate will save us a significant amount of time and effort on system scaling
management and thus allow us to allocate more workforce to feature development,” Ronald says.
Furthermore, autoscaling with serverless container deployment contributes to cost-effective operations for
both SHOPLINE and its customers. This method of deployment also ensures highly available websites with an
overall uptime of 99.9 percent, even during major shopping events.
Machine Learning as Next Phase of DevOps
Following its POS acquisition, SHOPLINE plans to continue its expansion into new countries and attract
larger enterprise customers. It also aims to introduce machine learning (ML) as the next phase of its DevOps
journey, to help merchants better plan for major sales events. ML models would be used for predicting site
traffic and autoscaling infrastructure on e-commerce sites before the sales begin and thus hedge risks from
anomalous website traffic.
With AWS, SHOPLINE is confident in its ability to deliver reliable e-commerce platforms to global customers.
“AWS solutions are highly available out of the box, from basic Amazon Elastic Compute Cloud [Amazon EC2]
instances to managed services like Amazon EKS and Amazon Elasticsearch Service. These solutions are a lot
more stable and cost-effective than building things on your own,” Ronald concludes.
To learn more, visit
aws.amazon.com/ecommerce-applications.
About SHOPLINE
SHOPLINE is a leading smart commerce platform in Asia, with offices in eight locations that have
helped more than 250,000 merchants (data up until 2021) open their online stores. SHOPLINE supports
brands of all sizes to achieve local and international growth with an omnichannel presence.
Benefits of AWS
- Helps detect and mitigate sophisticated cyberattacks
- Scales in less than 20 seconds instead of 3–4 minutes
- Ensures 99.9% availability across global regions
- Automates permissions and tedious maintenance tasks
- Implements container technology in 15 minutes instead of 1 month
- Encourages frequent innovation through an expanding range of new services