Project Overview
Paokinator is an AI-powered guessing game API server and client inspired by Akinator, where users think of an item (like an animal, object, or character) and the system asks yes/no questions to identify it through machine learning. It supports multiple domains beyond just animals, learns from user feedback to improve accuracy, and handles game state management with undo functionality.
Technologies Used
- Python: FastAPI for the REST API, NumPy/Pandas for ML computations, and a custom engine for prediction logic.
- Supabase: PostgreSQL database for data storage.
- Redis: Caching layer for performance.
- Docker: Containerization for deployment.
- Uvicorn/Gunicorn: Production server.
The project is designed for scalability with background sync and state persistence.
The Origin Story
The idea for Paokinator was conceived by Olisemeka Nmarkwe. The name “Paokinator” is a tribute to Dr. Kuo-Pao Yang, a respected professor at Southeastern Louisiana University. It’s a creative blend of his name and the game “Akinator,” reflecting the project’s academic and playful roots.
Our Team
This project was brought to life by a dedicated team: - Olisemeka Nmarkwe: Project Lead, Idea & Back-end - Angel Njoku: Team Member - Tyler Guidry: Team Member - Chris Walt: Frontend Help
Project Status & Commitment
Olisemeka is committed to keeping Paokinator running smoothly. You can play the game at https://paokinator.olisemeka.dev/. He will continue to host the website, moderate content to ensure a positive gaming experience, and refine the game’s algorithms to improve its accuracy.
While future versions of Paokinator are closed-source, Olisemeka is in the process of archiving and open-sourcing previous versions. This will allow anyone to use the older code for learning and experimentation. Thank you to all who have contributed in the past!
Screenshots
