Lovable Prompt Creator - AI Prompt Generator

## APPLICATION OVERVIEW
This application is a web-based tool designed to streamline customer service request triage using AI analysis. It assists support teams in classifying incoming requests, understanding urgency, and preparing initial responses, while ensuring human validation before final decisions.

## CORE FEATURES
1. **Request Input**: Customers submit their requests via a simple text box.
2. **Urgency Classification**: The AI analyzes the text to determine the urgency level based on keywords and sentiment.
3. **Categorization**: Requests are categorized into predefined categories for better organization and response handling.
4. **Initial Response Suggestion**: Provides a suggested initial response based on the request analysis.
5. **Human Review Workflow**: Ensures that all AI-generated classifications and responses are reviewed by a human operator before final submission.
6. **Analytics Dashboard**: Allows supervisors to view metrics on request handling efficiency and response time.

## DESIGN SPECIFICATIONS
- **Visual Style**: minimalist - Clean, simple design with plenty of white space, minimal color palette, and focus on typography
- **Color Mode**: Light theme with dark text on light backgrounds
- **Primary Color**: #1978E5 (accent for buttons, links, highlights)
- **Typography**: Use Inter from Google Fonts for headings, Inter for body text and UI elements
- **Border Radius**: 8px (moderately rounded) for buttons, cards, and inputs
- **Layout**: The main layout will include a top navigation bar, a central input area for request submission, an analytics dashboard section, and a sidebar for categorization options.

## TECHNICAL REQUIREMENTS
- **Framework**: React with TypeScript
- **Styling**: Tailwind CSS
- **UI Components**: shadcn/ui
- **State Management**: Redux for managing the state of requests and user sessions

## IMPLEMENTATION STEPS
1. **Set up the React project**: Initialize a new React project with TypeScript.
2. **Install necessary dependencies**: Install Tailwind CSS, shadcn/ui, and Redux.
3. **Create the main layout**: Design the main components including the navigation bar, request input area, and analytics dashboard.
4. **Implement the AI classification logic**: Integrate Google AI services for text classification and sentiment analysis.
5. **Build the request submission form**: Create a user-friendly form for submitting customer requests.
6. **Set up human review workflow**: Implement a review system that allows operators to validate AI-generated classifications and responses.
7. **Develop the analytics dashboard**: Create a dashboard for displaying metrics and analytics on request handling.
8. **Test the application**: Conduct thorough testing, including edge cases and user acceptance testing.

## USER EXPERIENCE
Users will interact with a clean interface where they can effortlessly submit customer requests. The AI will process these requests and provide classifications and suggestions in real-time, allowing support agents to focus on high-priority issues while ensuring a human touch in the final responses. The dashboard will provide insights into performance metrics, enabling continuous improvement in service delivery. 

### Additional Details
1. **Nome sugerido do aplicativo**: Classifica AI
2. **Problema que o app resolve**: Diminui o tempo de triagem de chamados e padroniza a classificação.
3. **Público-alvo**: Equipes de atendimento, suporte e supervisores de operação.
4. **Principais funcionalidades**: Entrada de solicitações, classificação de urgência, categorização, sugestão de resposta inicial, fluxo de revisão humana, dashboard de análises.
5. **Fluxo de uso passo a passo**: Usuário insere solicitação → AI classifica e sugere respostas → Operador revisa → Solicitação é concluída.
6. **Campos de entrada**: Texto da solicitação, urgência, categoria.
7. **Saídas esperadas**: Classificação, sugestão de resposta.
8. **Categorias e critérios de classificação**: Urgente, não urgente, reclamação, consulta.
9. **Prompt principal do modelo**: "Classifique a solicitação e sugira uma resposta inicial."
10. **Prompt de sistema ou instruções do modelo**: "Analise o texto da solicitação, determine a urgência e categorize a demanda."
11. **Exemplos de entrada e saída**: 
   - Entrada: "Meu pedido está atrasado há 8 dias."
   - Saída: "Classificação: Urgente; Sugestão de resposta: Pedimos desculpas pela demora e estamos investigando."
12. **Casos de teste**: Testar várias solicitações com diferentes níveis de urgência e categorias.
13. **Riscos e cuidados**: A IA pode cometer erros; sempre requer revisão humana.
14. **Melhorias futuras**: Integração com sistemas de CRM, aprendizado contínuo da IA com feedback dos operadores.

This comprehensive prompt should serve as a clear guide for the development of the application, aligning with the user's vision while adhering to best practices.