James Park (Northwind CTO) opened with an overview of their existing on-premise infrastructure: 47 microservices running on aging Oracle hardware, a legacy monolith handling 2.4M transactions/day, and three previous failed migration attempts with other vendors. The team carries significant technical debt from a 2018 platform rewrite that was never fully completed. Current hosting costs are $38K/month and climbing due to hardware refresh cycles.

Priya Sharma (VP of Compliance) outlined the non-negotiable regulatory constraints. Northwind operates in a regulated industry with strict data residency requirements — all production data must remain in US-East regions. Audit trails must be immutable and retained for 7 years. The compliance team needs to certify the new environment before any production cutover, which typically takes 4–6 weeks. Previous vendors failed because they underestimated these requirements.

Ben Carter presented the proposed three-phase migration approach: Phase 1 (Assessment & Planning, 6 weeks, $145K) covers dependency mapping, database profiling, and architecture design. Phase 2 (Migration & Implementation, 14 weeks, $380K) handles the actual containerization and data migration. Phase 3 (Optimization & Handoff, 4 weeks, $95K) focuses on performance tuning and operational documentation. Total engagement: $620K over 24 weeks.

Victor Huang (VP Engineering) raised concerns about zero-downtime migration given the 2.4M daily transaction volume. Ben outlined the blue-green deployment strategy with automated rollback triggers: if error rates exceed 0.1% or p99 latency crosses 500ms during cutover, the system automatically fails back to the on-premise environment. James noted this was the first time a vendor had proposed automated rollback criteria rather than relying on manual judgment calls.