AetherOps

Virtual MLOps Laboratory

Overview

AetherOps is a microservices-based web application that functions as a Virtual Laboratory for traditional Machine Learning models. It empowers users to tackle various machine learning problems (Regression, Classification, Clustering, and Dimensionality Reduction) through an interactive and modern interface.

Currently fully supporting Regression, the platform allows for real-time concurrent model training, dynamic hyperparameter configuration, and synthetic dataset generation.

🚀 Features

• Microservices Architecture: Completely containerized using Docker, with independent Backend and Frontend services.
• True Concurrency & Streaming: Powered by Python's ProcessPoolExecutor and Server-Sent Events (SSE), models are evaluated in true parallel. The frontend leaderboard streams and sorts the results in real-time.
• Dynamic Dataset Generation: Generate highly customizable synthetic datasets (e.g., number of samples, informative features, noise, test split).
• Extensive Model Support: Includes Linear Models (Ridge, Lasso, ElasticNet), Non-linear Models (SVR, KNeighbors), Tree Ensembles (RandomForest, GradientBoosting), and Neural Networks (MLP Regressor).
• Interactive UI/UX: Built with React and Vite. Features a premium "Glassmorphism" aesthetic, interactive particle backgrounds, Dark/Light mode toggles, and modal pop-ups for comparing hyperparameters.

💻 Tech Stack

Backend

• Python 3.11
• FastAPI (REST + SSE Streaming)
• scikit-learn (Machine Learning Core)
• asyncio & concurrent.futures

Frontend

• React 18 + Vite
• Vanilla CSS (Glassmorphism, CSS Animations)
• particles.js (Interactive backgrounds)

Infrastructure

• Docker & Docker Compose

🛠️ How to Run

1. Clone the repository.
2. Ensure you have Docker and Docker Compose installed.
3. Open a terminal in the project root directory and run:

docker-compose up --build

4. Once the containers are running, navigate to http://localhost:5173 in your browser.

📂 Project Structure

aether-ops/
├── backend/
│   ├── Dockerfile
│   ├── requirements.txt
│   ├── main.py          # FastAPI application & Streaming endpoint
│   ├── ml_core.py       # ML Pipeline, model factory, dataset gen & multiprocessing
│   └── schemas.py       # Pydantic data validation models
├── frontend/
│   ├── Dockerfile
│   ├── package.json
│   ├── index.html       # Vite entrypoint with particles.js CDN
│   ├── public/
│   │   └── favicon.svg  # Custom SVG Logo
│   └── src/
│       ├── App.jsx      # React Core Logic & Streams Reader
│       ├── index.css    # Global Styles & Themes
│       └── main.jsx
└── docker-compose.yml   # Multi-container orchestration

👨‍💻 Author

Built with ❤️ by Robert Lopez.