Enjoy playing around: https://pht5sypm5h.us-east-1.awsapprunner.com/

~# Video-Game-Advisor

Video-Game-Advisor is a containerizedweb application that leverages OpenAI’s GPT-3.5-turbo and the RAWG API to recommend video games based on user descriptions. With a retro gaming-inspired UI, the app displays personalized game recommendations complete with metadata (cover images, release dates, ratings, and links) in an engaging, responsive interface.

---

## Features

- **Personalized Recommendations:**  
  Generate 5 game recommendations based on a natural language description. If a game title is provided, it suggests that game and related titles.
- **API Integrations:**  
  Utilizes OpenAI’s GPT-3.5-turbo for natural language processing and the RAWG API for game metadata.
- **Modern UX:**  
  Responsive UI with retro gaming aesthetics, dynamic ticker animations, and social links.
- **Robust Error Handling:**  
  Implements the Post/Redirect/Get pattern to avoid duplicate submissions and provides friendly error messages when issues occur.
- **Containerized & Cloud-Ready:**  
  Packaged with Docker for consistent deployments, and integrated with AWS App Runner for scalable, 24/7 hosting.

## Tech Stack

- **Backend:** Python, Flask, Gunicorn  
- **APIs:** OpenAI GPT-3.5-turbo, RAWG API  
- **Containerization:** Docker  
- **Serverless Adapter:** Mangum (for AWS Lambda compatibility, if needed)  
- **Deployment:** AWS App Runner (with options for AWS Lambda/ECS)

## Getting Started

### Prerequisites

- [Docker](https://docs.docker.com/get-docker/)
- An AWS account (for deployment)
- API keys from [OpenAI](https://beta.openai.com/) and [RAWG](https://rawg.io/apidocs)

### Local Development

1. **Clone the Repository:**

   ```bash
   git clone https://github.com/yourusername/game-oracle.git
   cd game-oracle

2. Run Locally with Flask (for development):

   python app.py

   The app will be available at http://localhost:5000.

3. Run with Docker:

   Build the Docker image:

   docker build -t video-games-describer .

   Run the Docker container (assuming Gunicorn is configured to listen on port 8080):

   docker run -p 8080:8080 video-games-describer

   Then access http://localhost:8080.

Dockerfile

Your Dockerfile is set up to use the official Python 3.12-slim image, install dependencies, copy your application code, and start Gunicorn on port 8080:

# Use the official Python 3.12-slim image
FROM python:3.12-slim

WORKDIR /app

# Install dependencies
COPY requirements.txt .
RUN pip install --upgrade pip && pip install --no-cache-dir -r requirements.txt

# Copy application code
COPY . .

# Expose port 8080
EXPOSE 8080

# Run the app with Gunicorn
CMD ["gunicorn", "-b", "0.0.0.0:8080", "app:app"]

Deployment

AWS App Runner

1. Push to ECR:
   Tag and push your Docker image to an ECR repository in a supported region (e.g., US East (N. Virginia)):

   docker tag video-games-describer:latest 096771828335.dkr.ecr.us-east-1.amazonaws.com/video-games-describer:latest
   docker push 096771828335.dkr.ecr.us-east-1.amazonaws.com/video-games-describer:latest

2. Create an App Runner Service:

   • Log in to the AWS App Runner console.
   • Create a new service using “Container registry” as the source.
   • Select your ECR repository and image tag.
   • Configure service settings (instance configuration, environment variables, etc.).
   • Deploy your service and obtain the public HTTPS URL.

3. Custom Domain (Optional):
   Map your App Runner URL to a custom domain for a professional appearance.

AWS Lambda Deployment (Optional)

If you prefer a serverless approach using AWS Lambda, you can update your Dockerfile to use the AWS Lambda base image and integrate with Mangum. (See project documentation for details.)

Contributing

Contributions are welcome! Please fork the repository, create a feature branch, and submit a pull request with your changes. Follow best practices for code quality and testing.