High-volume CRM and Order Management provider came to Armakuni for full infrastructure migration from Origo DC and Rackspace to AWS EKS, hybrid cloud automation, database migration, and ongoing managed operations.
The challenge
A funded assessment sprint became a full infrastructure migration engagement that is now running as an ongoing managed operations retainer. Armakuni entered through an AWS-funded discovery phase, executed the full migration, and embedded a Flexible Engineering POD to run the platform on an ongoing basis.
Scale: 900+ data processing tasks on-premises, 5 independent VMs (FTP, web, and task servers), CRM platform serving e-commerce and direct marketing businesses globally.
What we built
Armakuni delivered full infrastructure migration from Origo DC and Rackspace to AWS EKS, hybrid cloud automation, database migration, and ongoing managed operations. The engagement ran as 3 SOWs: Task Migration Assessment (Phase 1 discovery and roadmap), Task Migration (Phase 2 execution), Flexible Engineering POD (Phase 3 ongoing managed operations). AWS funded the work at $19,800.
The outcomes
6 measurable outcomes shipped across the engagement. The ones that moved the business the most:
80-90% reduction in manual VM resource allocation. Karpenter and KEDA scale infrastructure automatically against real-time demand, so engineers focus on the product instead of provisioning and rebalancing.
Sub-5-minute auto-scaling response for database workloads. Demand spikes that once required manual intervention and degraded performance are now absorbed automatically, turning hours of engineering time into under five minutes with no one acting.
15-30% cost reduction across the infrastructure. Self-hosted MariaDB Galera clusters on EKS deliver full customer isolation far cheaper than separate managed instances, and right-sizing makes the platform cheaper than the on-premises estate it replaced.
100% of credentials moved into AWS Secrets Manager. Production passwords that had sat exposed in a code repository are gone, with every secret now encrypted, automatically rotated, and governed by role-based access with a full audit log.
All 900+ tasks and all 5 servers migrated and running in production on AWS. The entire on-premises estate across two third-party data centers has been moved, so their operational overhead and scaling constraints no longer exist.
Full real-time observability across the platform for the first time. The team went from learning about problems when customers called to catching them first, with every workload instrumented and every anomaly alerting.
Built on AWS
The production environment runs on 13 first-party AWS services. Delivered under our AWS Migration and Modernization, Cloud Native Development, DevOps Engineering competencies.
What's next
Ongoing platform optimization, cost governance, and continued Flexible Engineering POD retainer. Potential expansion of POD scope as the migrated platform scales.