AI suggestions showed in app, and follow ups

[Melestyl]

This pull request introduces a full AI suggestion and feedback workflow for ticket analysis, integrating database persistence, user feedback, and conversational refinement with the AI assistant. The changes span backend logic, database models and migrations, API enhancements, and internationalization for new UI features.

AI Suggestion & Feedback Workflow

• Added new Eloquent models AiSuggestion and AiFeedback to persist AI-generated solutions and user feedback, including relationships to tickets and users. [1] [2]
• Created database migrations for ai_suggestions and ai_feedbacks tables, storing model config, prompt hash, generated content, retrieved documents, and feedback actions.
• Added aiSuggestions relationship to the Ticket model for easy access to suggestions per ticket.

Backend AI Processing & Persistence

• Enhanced the AI backend (backend-ai/src/main.py) to save each suggestion to the database, including prompt hash, model config, and retrieved chunks. Added support for MySQL via pymysql and new environment variables for DB config. [1] [2] [3] [4]
• Modified the AI workflow to accept user feedback and previous suggestions, enabling conversational refinement. The prompt template now includes context from prior interactions and user corrections. [1] [2] [3] [4] [5]

Ticket Controller & API Enhancements

• Added aiFollowUp method to the ticket controller to validate feedback, retrieve latest suggestion, and push a refinement request to the AI queue, enabling users to request improved solutions.
• Updated ticket loading to include latest AI suggestions for display and further interaction.

Internationalization & UI Support

• Added new language strings in both English and French for the AI assistant interface, including summary, steps, feedback actions, and refinement prompts. [1] [2]

Minor Code Quality Improvements

• Minor formatting and code style improvements for better readability in ticket mapping logic.