✈️ FlightOnTime - Flight Delay Predictor

Executive summary of the project for the Flight Delay Hackathon. This system integrates Artificial Intelligence to predict whether a flight will be delayed based on historical and real-time data.

🚀 Project Architecture

The project is divided into four main modules:

1. 📂 Data Science (/Data-science)

• Objective: Exploratory Data Analysis (EDA) and model training.
• Models: Logistic Regression, Random Forest, XGBoost, and CatBoost.
• Stack: Python, Scikit-Learn, Pandas, Jupyter Notebooks.

2. ⚡ AI Microservice (/microservice)

• Objective: Serve predictions from the trained model.
• Tech Stack: FastAPI (Python), UV (package manager), Docker.
• Endpoints: Unit and batch predictions via REST API.

3. ☕ Backend (/backend)

• Objective: Data orchestration, user management, and prediction history.
• Tech Stack: Java 21, Spring Boot 4, Flyway, PostgreSQL (Docker).
• Documentation: Swagger/OpenAPI integration.

4. 🎨 Frontend (/frontend)

• Objective: Intuitive user interface for queries and metrics visualization.
• Tech Stack: Next.js 16 (App Router), React 19, Tailwind CSS 4, TanStack Query.
• Features: Prediction dashboard, interactive history, and AI assistant chat.

🛠️ Quick Installation

1. Requirements: Docker, Java 21, Python 3.11+, Node.js 20+.
2. Configuration:

   # Clone the repo and install frontend dependencies
   cd frontend && npm install
   # Start database and backend
   cd ../backend && ./mvnw spring-boot:run
   # Start AI microservice
   cd ../microservice && pip install -r requirements.txt && python main.py

📈 Impact

Optimizing passenger experience and airport operational management through accurate predictions based on airline, origin, destination, and distance.

---

Developed for the 2026 Flight Delay Hackathon.