When it comes to automating a deployment pipeline, Agile Stacks delivers. Just ask Matt Ferguson, the CTO of Zeeto.io. The powerful combination of AWS services and native Kubernetes gave Matt and his team the assurances of running enterprise AI workloads to meet Zeeto’s growing customer needs.
According to eMarketer research, 2019 will post likely numbers of worldwide digital ad spend of $333.25 billion. This translates to a lot of competition and new technology differentiation. Zeeto is not immune to these market forces. They too must continue to press forward with new innovations, speed, agility and differentiation. Faced with a rapidly-growing advertising platform and infrastructure constraints that threatened to limit Zeeto’s continued scalability, the Zeeto team made the decision to look for new infrastructure that could serve the organization’s needs.
Zeeto began down one of many paths to improving system performance. What seemed most logical was optimizing their compute infrastructure to meet the next generation AI-workload performance needed, no - demanded by their ever data-hungry consumers. They’d been using a Mesosphere-based implementation for many of its early benefits as a first mover for container orchestration. However, their team quickly discovered this would not be the case for too long as the open-source Kubernetes community began to work on competitor features. For example, DC/OS lacked critical features around networking, communication and management for developers. The lack of these features meant custom code had to written. And this made it tedious to maintain and improve upon. No auto scaling was a serious limitation to the company’s requirements.
“As an AI-driven digital advertising network, Zeeto needs a reliable, scalable and responsive infrastructure. That’s why we turned to Agile Stacks and AWS,” said Ferguson, who recently shared about his team’s experiences with Agile Stacks Kubernetes Stack.
Zeeto chose to implement Kubernetes with Agile Stacks and was impressed with the results. Infrastructure as code made it easier to automate, reduced human error and configured infrastructure more quickly and effectively. Auto-scaling improved efficiency and ensured that Zeeto was only using the resources they actually needed when they needed it.
Thanks to greater visibility of key performance metrics, Zeeto could see the savings Agile Stacks helped them achieve. Notably:
- 40% savings on their cloud bill by implementing a strategy with AWS Spot Instances
- 75% savings across infrastructure management operating cost, and
- A Kubernetes cluster is spread across three availability zones to reduce the impact of zone failures.
Zeeto’s developers no longer have the burden of infrastructure management. Agile Stacks provides them the necessary infrastructure at their fingertips to adhere to pipelines atop Kubernetes that they can trust. Infrastructure as code inherently offers reliability, scale, and security through proper technical implementation and process compliance. From a business perspective, Zeeto realizes greater cost savings and efficiency while strengthening end-user UX by leveraging the Agile Stacks platform.
“By adopting Agile Stacks on AWS, our DevOps team has been able to completely automate our software delivery process using an infrastructure-as-code approach. This has allowed us to meet the critical requirements we face in ensuring highly scalable and reliable deployments while continuing to innovate in a fast-paced, agile environment,” says Ferguson. Yet, the ultimate winner here are Zeeto's customers. This elasticity and agility of compute power can be leveraged by their smallest customers to their largest. There is no doubt having a customer-centric mindset put Zeeto down this path. There is a good lesson to learn - listen to your customers because maybe even they can innovate your product better and faster than you can.