Implement an Inference Controller for Nachet

## Context

With all the recent refactor of the Datastore, we implemented Controllers to ease maintainability and scalability of the projects' backend. However, we have not implemented any kind of Controller specific to Nachet. We would like to refactor the Nachet module and implement an InferenceController that accepts a `model` parameter containing the pydantic `Inspection` model. This controller will centralize the logic and allow better handling of the Inference through times for the Backend (just like Fertiscan's ``InspectionController``).

### Changes done in the past regarding controllers:
- Upgraded the User class to better manage a User and its various interaction
- Created a ContainerController to manage the container blob storage and keep its content relationship in the database
- Created an InferenceController to manage Fertiscan's main workflows

References: https://github.com/ai-cfia/ailab-datastore/pull/260

## Acceptance Criteria:
- Create an `InferenceController ` class within the module created with the inspection model as an input.
- Refactor existing functions (Update/Feedback, Delete, etc.) to be methods of the InferenceController.
- Ensure the Pydantic model for inspection is still utilized, but modify its usage to work within the InferenceController as a parameter.
- Maintain compatibility with the current picture classification process.
- Update documentation to reflect the changes in the module structure and provide examples of how to use the new InferenceController.