This project is the final version of a cross-platform mobile application developed to help users interact with the ticketing system of a hypothetical theater. The app integrates a chatbot that allows users to engage with the theater through a conversational interface, such as Messenger or similar chat apps. The chatbot enhances the user experience by providing essential information about shows, assisting with ticket bookings, and managing other requests in an intuitive and user-friendly manner.
The mobile application allows users to interact with the theater’s ticketing system for two different halls, each featuring distinct performances with afternoon and evening shows. The chatbot is powered by Retrieval-Augmented Generation (RAG) and function calling, allowing it to interpret user queries, retrieve relevant information, and respond accurately based on real-time data from the theater’s system.
For actions like booking tickets, canceling reservations, or submitting complaints, the chatbot also provides quick action buttons, helping users seamlessly navigate to the necessary pages or forms.
The theater ensures accessibility by offering special accommodations for hearing-impaired users, including seats near interpreters and shows with supertitles.
- Real-time Show Information: Users can ask the chatbot for information about current performances, including which hall they are being shown in and the schedule for afternoon or evening showings.
- Ticket Booking: The application walks users through booking tickets for specific shows, allowing them to select seats, confirm the reservation, or cancel before finalizing.
- Ticket Cancellation: Users can cancel previously booked tickets directly through the chatbot in a few simple steps.
- Complaints and Feedback: The chatbot assists users in submitting complaints or feedback and can escalate issues to human agents when necessary.
- Accessibility Support: The app includes seating near interpreters for hearing-impaired users and provides supertitles for certain shows, enhancing the accessibility of performances.
- Error Handling: The chatbot uses confirmation dialogs to ensure that no unwanted actions are taken. If the system detects repeated misunderstandings, it escalates to a human agent to assist the user.
The chatbot utilizes intent recognition to understand user requests in natural language. Based on the detected intent (e.g., checking show times, booking a ticket, or submitting a complaint), it retrieves the appropriate information and responds accordingly. Additionally, the system can trigger function calls for specific actions like booking a seat or handling a complaint.
The communication between the chatbot’s Large Language Model (LLM) and the app’s frontend is facilitated by a Flask server, which manages requests and responses to ensure smooth operation.
Apart from the chatbot, the application includes a fully functional backend system that enables users to:
- Book tickets
- Cancel tickets
- Access information on shows
- Submit complaints
- Get information about the theater
These operations can be completed both through the chatbot and directly via the app interface, offering users multiple interaction options.
- Flutter (Dart): Cross-platform framework used to build the mobile app, making it available on both iOS and Android devices.
- Flask (Python): Backend framework that manages the communication between the mobile app’s frontend and the chatbot server, handling queries and responses.
- Large Language Model (LLM): Utilized for natural language processing, enabling the chatbot to understand user queries and identify user intent.
- Retrieval-Augmented Generation (RAG): Ensures the chatbot retrieves relevant information from the theater's backend system based on user queries, enhancing real-time information delivery.
- Clone the repository from GitHub:
git clone https://github.com/your-repo/theater-chatbot.git
- Navigate to the lib/ai_script directory:
cd lib/ai_script - Run the cells of the mistral-chat-with-rag.ipynb notebook to instantiate the server.
- Connect your mobile phone to your development environment or use an emulator to start the Flutter app:
flutter run
- You are all set!
Check out the demo video to see the app and chatbot in action, showcasing their key features and how they provide a smooth, mobile-friendly user experience.