Case Study
Atlas Air — Governing a 1,000-Server AWS Migration Through Acquisition
Atlas Air
Role
Fractional VP of Cloud Architecture & AI Integration
Period
October 2025 – March 2026
Duration
6 months
Reporting Line
VP of IT and Director of IT
Industry
Air Cargo / Aviation
The Situation
Atlas Air — one of the world's largest air cargo operators — had been acquired, and the new owner mandated a full migration of all on-premises infrastructure to AWS within a hard deadline. Data center leases were expiring; there was no flexibility on the timeline.
The complication: Atlas Air had no internal cloud personnel. Accenture had managed their on-prem infrastructure for years. Rackspace had been brought in by the investors to execute the migration. But after almost a year of effort, the program was spinning its wheels — even getting an AWS Landing Zone up and running had taken longer than anyone expected.
There was no senior internal authority holding the vendors accountable, enforcing AWS Well-Architected discipline, or owning the migration architecture end-to-end.
Why a Fractional Executive
A 1,000+ server migration on a fixed deadline, executed across multiple vendor teams (Accenture, Rackspace, internal IT, internal security), with simultaneous AI governance work for post-migration Bedrock adoption — this wasn't a permanent VP role Atlas Air could afford to wait six months to fill. They needed credible cloud architecture leadership in the room immediately, with the seniority to hold vendors to standard and the architectural depth to make the right calls in real time.
What I Led
- •
Vendor governance and architectural authority. Served as the primary architecture and delivery authority across Accenture, Rackspace, internal engineering, infrastructure, and security teams. Held scope and schedule across vendor boundaries throughout the program.
- •
The migration program itself. Sequenced and governed migration waves for 1,000+ servers — on schedule and within scope. Designed disaster recovery with multi-region automated failover.
- •
AI governance framework. Designed the multi-account Bedrock governance model with least-privilege IAM, model invocation logging, and cost controls — establishing patterns for future RAG pipelines integrating enterprise data sources with Bedrock Knowledge Bases.
- •
Network and security architecture. Designed the global multi-account environment with AWS Cloud WAN, Direct Connect, Cisco DMVPN, Palo Alto VM-Series firewalls, and centralized inspection VPCs. Built RBAC integrated with CyberArk for privileged access. All deployed via CDK and Terraform.
- •
Cost engineering. Architected storage modernization using FSx for NetApp ONTAP delivering $30K+ in projected annual savings. Rationalized workload placement for long-term FinOps efficiency.
- •
Application modernization design. Designed serverless data pipelines (Lambda, Step Functions, EventBridge) for operational integration. Designed migration strategies for Oracle RAC, AIMS, and high-availability Globalscape SFTP clusters.
- •
Executive reporting cadence. Established architecture review and risk management cadences enabling cross-vendor accountability and executive-level visibility throughout the migration lifecycle.
Business Outcomes
- •
1,000+ servers migrated to AWS under tight executive oversight, on schedule
- •
Multi-vendor delivery aligned to a single architectural direction
- •
AI/GenAI adoption path that was governed and de-risked before scale
- •
$30K+ projected annual savings through storage modernization
- •
Comprehensive disaster recovery posture with multi-region automated failover
What I Left Behind
- •
A production-ready, multi-account AWS footprint
- •
A documented AI governance framework ready for Bedrock adoption
- •
Architecture patterns for future RAG pipelines and enterprise data integration
- •
An internal team aligned on the next phase of adoption — with my responsibilities cleanly handed off across multiple individuals