Adaptive Learning Path Generator with TypeScript

Design a type-safe TypeScript microservice that dynamically generates personalized student learning paths using machine learning algorithms. Create an architecture that can ingest student performance data, utilize generics to handle multiple curriculum types, and generate adaptive progression recommendations. Implement robust error handling for data inconsistencies and include performance tracking decorators. The solution should support multiple educational domains and handle complex curriculum mapping.