Evaluation: Restructuring code

[AkhileshNegi]

Is your feature request related to a problem?

The evaluation feature code doesn't follow the standard Routes → CRUD → Core architecture used in the rest of the codebase. This makes it harder to maintain, understand where new code belongs, and onboard developers to this feature.

Describe the solution you'd like

Refactor evaluation code to match the standard pattern:

• Split oversized evaluation.py routes file (731 lines) into focused modules
• Reorganize CRUD layer
• Move complex orchestration logic to a service layer (app/services/evaluation/)
• Consolidate all public CRUD functions in init.py for consistent imports