The challenge
A telecommunications operator needed to move from a legacy Windows-based platform to a Linux-based product on AWS. The incompatible data models and limited sizing benchmarks made the migration more than a simple infrastructure move.
My contribution
I worked with a database specialist to map legacy records to the new platform, wrote Bash and REST API tooling to transform and import customer data, and personally executed the production data migration. When production traffic exposed scaling limits after cutover, I analyzed application behavior, logs and AWS resource data with the engineering team.
Outcome
The team restored stable service in approximately 24 hours. After demand stabilized, frontend capacity was reduced from 13 servers to 10, a 23% reduction from peak, while six backends remained. The outcome was a working production platform and lower frontend resource requirements, not a claimed 23% reduction in the total AWS bill.
Engineering perspective
Validate the behavior of the application at production scale, not merely the health of its infrastructure. Cloud hosting does not automatically make an application cloud-native.