Pratheek22/Distributed-ML-Orchestration-Platform ? reverse-engineered prompt

Build me a simple web app for distributed ensemble machine learning, where someone can sign up, log in, upload a CSV dataset, choose the target column, pick a model like decision tree, random forest, or logistic regression, and start training from a dashboard.

I want the training to run across multiple worker nodes, with a central master that splits the data, sends work out asynchronously, and combines the results when it’s done. Please make it handle workers joining and failing gracefully, show job status and active workers in the UI, and let the user download the trained model once the job finishes.

Keep the design clean and organized, with a backend service for orchestration, separate worker services for training, and a lightweight frontend for login, upload, job setup, and monitoring. Use MySQL for users, datasets, and jobs, plus messaging and worker discovery so the system can scale and recover if a worker goes away.