You are leading a complex data migration from a legacy operational data store into a modern analytics and operations platform. The migration matters because downstream workflows, reporting, and operational decisions will move onto the new environment, but the source system has inconsistent schemas, undocumented business logic, and years of data quality issues. Some teams want a fast cutover to unlock new capabilities, while operations leaders are more concerned about continuity and auditability. You also know that data consumers have built manual workarounds around the legacy system, so hidden dependencies are likely.
Walk me through how you would manage this data migration project. What are the key stages and potential failure points?