When a construction crew severed the primary fiber line for Carter Credit Union in northwest Louisiana, an aging on-premises infrastructure left the $770 million institution completely offline for 24 hours, exposing the severe risks of legacy data center dependencies and sparking a strategic migration to Amazon Web Services.
The Bottom Line
- Carter Credit Union executed a full migration from an on-premises VMware stack to AWS utilizing a lean, three-person internal IT team without hiring external consultants.
- The transition is projected to save the institution more than $70,000 over a three-year period by eliminating recurring hardware replacement cycles and VMware licensing renewals.
- To reach members in rural and underserved areas without traditional branches, the institution deployed CarterLIVE! Interactive Teller Machines, backed by a hybrid cloud model splitting workloads between AWS and the vendor platform.
Infrastructure Vulnerabilities at a 70-Year-Old Credit Union
Founded in 1954 for International Paper company employees in Springhill, Louisiana, Carter has grown into an institution serving more than 45,000 members across 13 locations. For decades, the credit union relied on a traditional VMware virtualization stack hosted on physical servers in a single data center, tethered to a solitary fiber circuit. That architecture proved fragile. When the primary fiber line was cut by construction crews, a backup circuit routed through the exact same conduit failed simultaneously, knocking the organization offline for a full day. A subsequent hardware failure triggered a multiday outage that left critical systems unavailable for 3 consecutive days.
NEW: A Carter Credit Union class action settlement offers cash and credit monitoring to individuals whose info was compromised in a 2025 data breach.https://t.co/8VBZj3Pqtv
— https://x.com/ClassAction_org/status/2103590435939566023
We had a single point of failure in almost every layer of our infrastructure,
said Jason Berard, Carter’s chief information technology officer.
Although the team had relocated some workloads to a colocation facility in Dallas, Texas, the arrangement still forced a small staff to manage physical hardware, keep up with firmware updates, and budget for costly server replacement cycles every 3–5 years. The total cost of ownership climbed steadily while member-facing outage risks remained unacceptably high.
A Phased Cloud Migration Managed by Three People
In 2022, Berard evaluated options across major cloud providers, including AWS, Google Cloud, and Microsoft Azure. The deciding push came from a peer recommendation: the president of a credit union service organization shared a successful cloud transition story. Rather than attempting a risky, all-at-once lift-and-shift over a single weekend, Carter adopted a measured, phased strategy led by Berard and two infrastructure specialists.
The team started by targeting servers that were already approaching required operating system upgrades. Instead of spending capital on aging hardware and renewing VMware licenses, they rebuilt those workloads directly in AWS. As the process matured, they utilized AWS Transform MGN to streamline the migration of remaining systems. To bridge any skill gaps, team member Randy and his colleagues completed targeted AWS training focused on cloud server management, auto scaling, and cost optimization, allowing the credit union to modernize its core operations entirely in-house.
Financial Savings and Modernized Service Delivery
By moving to a consumption-based cloud model, Carter pays only for the compute and storage resources it actually consumes, avoiding the financial trap of over-provisioning for peak capacity. Over a 3-year analysis period, the credit union is projected to save more than $70,000 by bypassing hardware refreshes and software licensing fees. Those funds are redirected toward member services, financial education, and community development initiatives.
The operational shift also overhauled how the cooperative serves rural communities. Carter deployed an Interactive Teller Machine network branded as CarterLIVE!, connecting members to live tellers via two-way video alongside standard ATM capabilities. The institution is now migrating its ITM infrastructure to a hybrid cloud model, splitting components between AWS and the vendor’s cloud platform to match the robust redundancy achieved across the rest of its modernized ecosystem.
| Metric / Factor | Legacy On-Premises Architecture | AWS Cloud Infrastructure |
|---|---|---|
| Primary Infrastructure Risk | Single points of failure across servers and fiber links | High availability and automated redundancy |
| Hardware Refresh Cycle | Required every 3–5 years | Eliminated via consumption-based cloud scaling |
| Compliance and Security | High overhead for local monitoring and controls | Built-in financial regulatory frameworks and encryption |
| Team Allocation | Strained internal staff managing physical hardware | Streamlined operations managed by a three-person team |
Disclaimer: The information provided in this article is for educational and informational purposes only and does not constitute financial advice.
