Scope: Development of Automated Customer Service Chatbot for Publi

1. Project Objective Develop an automated customer service chatbot aimed at facilitating the submission and tracking of service requests in various public service categories. The chatbot will use artificial intelligence (AI) to interpret and respond to user inquiries without the need for pre-configured flows. It will operate based on a provided instruction document and will integrate with communication channels such as WhatsApp and Instagram (via Meta's official APIs), along with a control panel for managing and tracking service requests. 2. Main Features- WhatsApp API Integration: Implement the chatbot to serve and respond to citizens via the official WhatsApp API by Meta.- Instagram API Integration: The chatbot must also be available through Instagram's official API, allowing citizens  to submit queries and open service requests directly through Instagram Direct Message (DM).- Advanced AI: Utilise AI based on ChatGPT or a similar solution to understand and respond to user inquiries by  interpreting the service instructions document.- Service Request Submission: The chatbot must enable citizens to submit service requests in different public service  categories by providing the necessary information and categorising the request accordingly.- Control Panel (Helpdesk System):

  - Service Category Management: Create a panel or website where the public organisation can define the categories    of service requests citizens can open (e.g., Public lighting, healthcare, sanitation, etc.).  - Viewing and Tracking Requests: The panel will allow department managers to view open requests, filter by category,    and track the progress of each service request.  - Status Updates: Through the panel, department managers will be able to update the status of the service request    (e.g., "In Progress", "Completed", "Pending"), as well as add internal comments or responses to the citizen.  - Reports and Statistics: The panel should offer features for generating reports and performance statistics (number    of requests, resolution time, most requested categories, etc.).- Service Instructions Documents: The chatbot will read and interpret pre-defined documents (in pdf, docx, or similar  formats) containing instructions for handling various service requests.- Customisation per Public Organisation: The system should be customisable for each public organisation, allowing  for personalised service categories, request workflows, and specific service configurations.- Service Request History: Citizens should be able to check the status and history of their previous service requests  through the chatbot. 3. Technical Requirements- Required Knowledge:  - Artificial Intelligence and ChatGPT: Implementation and fine-tuning of AI models based on ChatGPT (OpenAI)
or similar solutions.  - WhatsApp and Instagram API: Experience with the official Meta API for integration and message handling    via WhatsApp and Instagram Direct Message.  - Backend Development: Creation and management of a backend to handle service requests and integrate the services,    along with a web interface for the control panel.  - Document Integration: Ability to extract information from documents in various formats (pdf, docx, etc.) For AI interpretation.- Technologies:  - Programming languages: Python, Node.js, or other suitable languages for ai and api integration.  - AI frameworks: OpenAI GPT, TensorFlow, or similar.  - Integration with databases (MongoDB, MySQL, or PostgreSQL).  - Familiarity with RESTful APIs.  - Development of responsive web interfaces for the control panel. 4. Additional Features (Desirable)- Multi-platform: In addition to WhatsApp and Instagram, the chatbot should also function on other communication  channels (such as Telegram, Webchat, etc.).- Advanced Reporting: Automated, detailed reports on performance, most attended categories, and other key indicators.- Automated Notifications: The system should send automated notifications to the citizen with updates on the  service request status.
5. Delivery Timeline- Estimated Timeline: 3 to 5 weeks (negotiable based on complexity).- Partial Deliveries:  - First Delivery: Basic chatbot functionality with service request submission (1-2 weeks).  - Second Delivery: AI integration and automated responses (2-3 weeks).  - Third Delivery: Full integration with WhatsApp, Instagram, and control panel, with final testing (4-5 weeks).  - Testing and Adjustments Phase: Final phase for correcting and adjusting functionalities. 6. Documentation and Support- Documentation: Complete documentation of the code, integration instructions, API usage guide, and user manuals  for the managers, as well as the control panel operation manual.- Support: Ensure support during the implementation and post-implementation period (to be negotiated). 7. Required Qualifications- Proven experience in developing chatbots.- Practical knowledge of ChatGPT or similar AI solutions.- Experience with WhatsApp and Instagram API (official Meta APIs).- Experience in developing helpdesk systems or control panels for managing workflows.- Ability to deliver a scalable and secure system. 8. Language Support- The system must also support Portuguese (pt-BR) to allow for seamless communication with users in Brazil.