[spike] Enable IA to decision on provision or extend mac machines

[adrianriobo]

Get data information

• Record each request (team, time of execution, UTC datetime)
• Record costs
• Record unattended requests / Waiting time

Feet an IA model:

• Linear or Polynomial Regression

This type of model could be useful if you're looking to predict a continuous variable, such as the number of machines needed or the duration of requests. You can adjust the model to minimize costs or maximize performance.

• Time Series Models:

If your data has a strong temporal component, such as daily or seasonal patterns, you might consider time series models like ARIMA (Autoregressive Integrated Moving Average) or more advanced models like LSTM (Long Short-Term Memory).

• Logistic Regression or Classification:

If you are more interested in making binary decisions, such as whether to provision a new machine at a given time, you could use logistic regression for binary classification problems.

Deploy the model and handle the shared pool based on it.