Zé Delivery Builds Beverage Delivery Service on AWS, Supports Exponential Growth
As more people than ever pick up their phones to order beer and other drinks for home delivery, Zé Delivery, a subsidiary of international beverage producer Anheuser-Busch InBev Companies (AB InBev), built a service that delivers cold beverages to customers quickly, wherever they are. Using Zé Delivery’s app and website, customers can connect to an extensive network of beverage distributors in Brazil. This direct-to-consumer strategy is changing the way consumers and distributors buy and sell beer, and Zé Delivery is at the center of the trend.
Zé Delivery was facing a challenge—its single database couldn’t handle the exponential growth that the company experienced in 2020 as consumer behavior shifted wildly in response to the COVID-19 pandemic. The company’s database resources were not elastic enough to cost-effectively manage the regular spikes in demand that accompanied weekends, Wednesday night soccer matches, and other occasions. By migrating its data persistence infrastructure from a monolithic architecture to a more agile, elastic cloud-native infrastructure on Amazon Web Services (AWS), Zé Delivery could continue to deliver a quality user experience while sustaining hypergrowth.
Enabling Elasticity in the Cloud
Zé Delivery designed its early data infrastructure to be simple and stable, but the company’s needs changed quickly as it grew. It has been using AWS as support infrastructure since 2016. At that point, the company had only three engineers using AWS tools. “We saw that by relying on AWS, we could do more with fewer people,” says Marcelo Malcher, chief technology officer of Zé Delivery. In 2019, the company expanded its use of AWS, migrating its existing database to Amazon Aurora, a fast relational database built for the cloud. “That let us spread the read load to up to 16 synchronized database instances when needed, so we could handle a higher number of orders and requests,” says João Filipe Storarri, senior backend engineer at Zé Delivery.
Since the first iteration of its serverless architecture, the company’s small engineering team has used a combination of AWS Lambda, a serverless compute solution, and Amazon API Gateway, a fully managed secure API service, to iterate quickly and sustain growth. Then, at re:Invent 2019, AWS announced that AWS Hyperplane, a network function virtualization service, would make it simpler and faster to provision internet protocol addresses and connect to Amazon Virtual Private Cloud (Amazon VPC) networks—isolated virtual networks for launching AWS resources. Zé Delivery realized this technology would enable it to scale even further.
AWS provides all the technologies and services we need to evolve our online solution. It enables us to rapidly scale our infrastructure, and it’s helping us improve the quality of our service as we grow.”
Chief Technology Officer, Zé Delivery
Expanding Cloud Capabilities to Meet Skyrocketing Demand
“At the beginning of 2019, we were delivering around 20,000 monthly orders,” says Malcher. “By December, we were processing 350,000 monthly orders. That was an amazing year for us.” The company expected its steady progress to continue during 2020. Instead, the number of users and orders skyrocketed. “Everyone was staying home and relying on delivery services like Zé Delivery,” says Malcher. “We had huge growth. That April, we surpassed 1 million monthly orders.” In December 2020, Zé Delivery processed 4.5 million orders.
Zé Delivery’s total annual orders increased from 1.6 million in 2019 to 27 million in 2020—a 1,688 percent growth year over year. “Without the work on Aurora that we’d been doing since 2019, we wouldn’t have been able to handle that growth,” says Malcher. By putting a proxy layer on top of its data, Zé Delivery optimized how its application routes each database request so that the company could manage the influx of users. Using serverless architecture on AWS Lambda also enabled Zé Delivery to quickly experiment with using AWS services and better respond to evolving needs.
Using a Serverless Architecture to Offer an Industry-Leading User Experience
Moving to AWS Lambda has also improved the company’s ability to handle its data throughput and optimize its delivery chain part by part. As its user base grows, Zé Delivery is processing an increasing amount of geographic data to identify areas where the quality of service needs improvement. “We’re using AWS to understand what we can do to provide the best experience to our users,” says Malcher. “We’re quickly developing pipelines to help our business team make the best decisions about what to spend and where to focus efforts.” AWS Lambda enables the company to call on as much compute power as it needs when under load and to minimize server capacity when idle.
Using AWS Lambda also gives Zé Delivery breathing room to continue to grow its team. The company works in “squads”—focused teams working on specific areas. “When we had only eight people, we worked in one or two squads,” says Malcher. “Now we have 30 squads. We’re working in separate domains, breaking our monolithic application into smaller services while still relying on AWS Lambda for the compute layer.” Since the beginning of 2020, Zé Delivery has grown from 60 to 300 employees.
The company takes advantage of its serverless foundation to continuously adjust the services it uses, avoid traditional infrastructure costs, and speed up innovation. “Using AWS Lambda and Amazon API Gateway together makes fast iteration possible. We can make several deployments per day,” says Storarri. “AWS provides all the technologies and services we need to evolve our online solution,” says Malcher. “It enables us to rapidly scale our infrastructure, and it’s helping us improve the quality of our service as we grow.”
Zé Delivery prioritizes quality of service above rapid expansion, even as its user base has grown by 10 times year over year. “If we were growing but not offering the best experience for our customers, we’d have to stop expanding to address the issue,” says Malcher. Reliability is another crucial component to Zé Delivery’s success. “AWS has so many solutions, and reliability is in its bones,” says Storarri. Zé Delivery’s constant pursuit of delivering a better customer experience has led to higher ratings and has helped it become a leading beverage delivery service in Brazil.
Delivering Innovation in a Bright Future
Zé Delivery continues to make the most of AWS services by optimizing its architecture to be even more agile and efficient. “AWS provides us with the flexibility to figure out which solution will help us best serve our customers,” says Malcher.
To help it add powerful functionality to its app, the company is currently developing a machine learning layer powered by Amazon SageMaker, a service that enables users to prepare, build, train, and deploy high-quality machine learning models quickly. “We’re planning to use machine learning on AWS to do cross-selling and offer recommendations for our users,” says Malcher. “We want to use it to improve our logistics.” Machine learning will help Zé Delivery compare and optimize the performance of each step in its delivery system.
For Zé Delivery, the key to innovation is agility. “There is no perfect architecture; you’re always trying to choose the right technology for the right time,” says Malcher. “AWS has the technologies and services we need to continue to evolve.”
About Zé Delivery
Zé Delivery is a startup beverage delivery service operating in Brazil. As a subsidiary of international beverage producer AB InBev, the startup received support from ZX Ventures, which has a mandate from AB InBev to invest in innovative products across industries.
Benefits of AWS
- Grew its user base by over 10x year over year
- Increased from 1.6 million to 27 million annual orders
- Enabled real-time access to data and analytics
- Optimized performance by region
- Grew its team from 60 to 300 employees
AWS Services Used
AWS Lambda is a serverless compute service that lets you run code without provisioning or managing servers, creating workload-aware cluster scaling logic, maintaining event integrations, or managing runtimes.
Amazon API Gateway
Amazon API Gateway is a fully managed service that makes it easy for developers to create, publish, maintain, monitor, and secure APIs at any scale.
Amazon Aurora is a MySQL and PostgreSQL-compatible relational database built for the cloud, that combines the performance and availability of traditional enterprise databases with the simplicity and cost-effectiveness of open source databases.
Amazon SageMaker helps data scientists and developers to prepare, build, train, and deploy high-quality machine learning (ML) models quickly by bringing together a broad set of capabilities purpose-built for ML.
Companies of all sizes across all industries are transforming their businesses every day using AWS. Contact our experts and start your own AWS Cloud journey today.