The docker build should be able to run the trading bot python script in a single container, while running as a celery task and handled by the environment's cron job (or cron tab). Separately, it should be able to communicate with a docker container running the mt5 running on wine (for easier deployment to any linux VPS/cloud/kubernetes cluster/etc)

• Celery Task
• Functioning Docker build of the Python script
• Scheduled and managed by the cron job (either env or by cron tab)
• Be flexible enough to direct its mt5.initialize's path to an executable running in a separate container

Noticing the slow performance of Python, Cython can process much faster, hence, being able to process at smaller/lower time-frames much more predictably.

Whichever the cloud platform is decided, it should be automatically deployed after each enhancement change is applied, provided that #2 is complete to ensure stability and confidence is maintained.

As part of the experiment, all tests, training, fitting and learning how machine learning identifies patterns and predicts the next closing price, it gets all stored in a Jupyter Notebook where it can be reproduced and further researching each model and approach to solving market data behavior and movements. Would it also make sense to document these in part of the trading system's development in a wiki style?

For local development only, train the model and return a standard algorithm report on, but not limited to:

• % wins vs % loss
• number of wins vs number of loss
• number of long vs short trades
• mean square error
• predictable chart vs actual chart vs test chart

Have to make sure that the code continues to work as its developed with new features. Unit tests of the python script and the docker script is required.