An AI-powered marketing platform that helps brands automate campaigns, improve targeting, and scale creative output across digital channels came to Armakuni for structured cloud cost control and operational visibility across its containerized AWS platform.
The challenge
Pixis set out to scale its AI marketing platform while taking tighter control of cloud cost and operational risk. Manual coordination across 50+ interdependent microservices slowed releases, nine distributed AWS accounts blurred billing and governance, and limited real-time insight into data transfer, EKS workloads, backups, and storage pushed infrastructure cost up with no container-level attribution to show which services were responsible.
What we built
Armakuni delivered structured cloud cost control and operational visibility across its containerized AWS platform. The engagement ran as a cloud operations and FinOps engagement covering Amazon EKS standardization with GitOps pipelines, AWS account realignment, and AWS CUDOS cost-visibility dashboards.
The outcomes
4 measurable outcomes shipped across the engagement. The ones that moved the business the most:
25% reduction in overall cloud spend. The platform scales its AI marketing workloads on AWS without infrastructure cost rising at the same pace.
48% compute cost savings. Right-sizing EKS workloads to real usage keeps recurring compute cost tracking actual demand rather than worst-case provisioning.
99.996% platform uptime maintained through the migration and cost optimisation.
4 structured AWS accounts, consolidated from 9. Clear cost attribution makes ownership, billing, and governance across the platform unambiguous.
Built on AWS
The production environment runs on 11 first-party AWS services. Delivered under our AWS Migration and Modernization Consulting, DevOps Consulting competencies.
What's next
Extending CUDOS-driven FinOps into automated right-sizing and forecasting, and broadening anomaly detection and budget alerting across the platform.
Architecture
