The Challenge
A national telecommunications provider had committed $150M to a cloud migration initiative. Six months in, the program was stalling: teams were migrating applications that didn’t need to move, skipping ones that did, and the infrastructure architecture was incompatible with the AI workloads leadership had planned for Phase 2.
Our Approach
We deployed BenchMark to provide a data-driven foundation for the migration program. We executed the migration across three areas: assessing the tech stack to score 340 applications for cloud compatibility, benchmarking six model architectures against telecom workloads, and generating an infrastructure blueprint to optimize compute provisioning.
The Results
- $45M saved in misallocated spend: Stopped unnecessary migrations and right-sized infrastructure provisioning.
- 340 applications scored and prioritized: Objective, data-driven migration sequencing replaced subjective prioritization.
- 60% faster migration execution: Clear priorities and right-sized infrastructure eliminated the false starts and rework that had plagued the first 6 months.
- AI-ready from day one: The new infrastructure blueprint natively supported the AI workloads planned for Phase 2, avoiding a costly re-architecture.