Case Study
Refactoring cloud architecture to reduce costs
Client
ReMasterMedial is a platform for digital audio mastering. It is designed for podcasters, musicians, content creators, and broadcasters to improve the quality of their audio and video content and prepare it for specific media distribution. The platform emphasizes simplicity, speed, and quality and helps thousands of users worldwide achieve professional audio levels without the need for complex audio post-processing in the studio.
Challenge
The objective of the project was to modernize both the application architecture and the underlying infrastructure of the ReMasterMedia platform. Key priorities included transitioning to container run-time, adopting modern cloud-native technologies, and optimizing for cost efficiency — all while preserving the platform’s high standards for quality and performance. Given that the platform processes tens of thousands of minutes of audio and video content each month, the solution required a strong focus on scalability, high availability, and computational efficiency.
Solution
The original architecture relied on components developed in high-level programming languages to process audio and video files and live streams. This was replaced with a low-level implementation embedded directly into the open-source FFmpeg framework, resulting in a significant reduction in system resource consumption for signal processing. The entire platform was containerized and migrated from virtual machines on AWS EC2 to a serverless architecture using AWS Fargate within the Amazon ECS service, with autoscaling configured to meet varying demand. Additional cost and performance optimizations included transitioning from x86 CPUs to energy-efficient ARM64 architecture. Infrastructure management was fully automated using Infrastructure as Code (IaC), replacing manual configurations to ensure consistency, repeatability, and simplified environment management. As a forward-looking step, a proof-of-concept was conducted to evaluate AWS Lambda as a potential future runtime environment for signal processing workloads.
Results
The redesigned architecture achieved a 70% reduction in AWS operational costs while maintaining the platform’s reliability. Minor performance trade-offs were observed at low traffic levels but remained virtually unnoticeable to end users. Thanks to autoscaling, the system now responds more efficiently to peak demand. Implementing Infrastructure as Code streamlined environment management, eliminated manual processes, and enabled faster, safer infrastructure changes. The success of the modernization effort also initiated plans to extend similar enhancements to other parts of the platform, where comparable efficiency gains and cost savings are anticipated.
"Through architectural refactoring and infrastructure optimization, we significantly reduced AWS costs while maintaining the high standards of quality and reliability our users expect. Zonio’s deep technical expertise and proven experience with modern cloud technologies provided the confidence needed to pursue and implement these transformative changes."
Andrew Knox
Founder & CEO
Technologies Used
AWS - ECS, Fargate, Cloudwatch, FFmpeg, OpenTofu
Are you considering modernizing your applications or infrastructure?
A well-architected system not only streamlines development and maintenance but can also lead to substantial reductions in operational costs. We help identify targeted opportunities where modern technologies can drive both cost efficiency and performance improvements.