Machine learning / Data science / Data analytics?

Question

Neural networks

Good day! There is a task (or rather Wishlist to which I want to come)!
There is a full-time aggregator that pulls up information from a variety of APIs (about 60+), the subject is the same - everything related to the construction / repair of buildings, and production, logistics + banks!
Every day, half of the data for small products is updated steadily, and data for large buildings such as a multi-storey building is updated every 3 months.
To understand further what I need, I will describe an approximate process:
An order for the construction of a plot for farmsteads comes -> estimates, managers of 2-3 days surf our aggregator and select suitable suppliers (choose according to the offered materials, etc.) -> then they choose a suitable logistics company at the moment (almost every time different, but periodically return to the old ones) -> then there are phone calls, deals, etc.
Wishlist is to automate the process of selecting suppliers and logistics.
There is actually a lot of data, but they are not processed in any way.
The result should be a recommendation system using an example: We will win by ordering from this supplier in terms of price, but we will lose in terms of delivery; We will win if we order from several suppliers, but we will use several logistics companies, etc. In the flesh to the worst decision!
The question is which of these to use or all together Machine learning / Data science / Data analytics ?
In this I am a complete zero, (understanding at the level of ML = SkyNet (sarcasm)).
I want to spend all my free time studying this subject area, but I don’t know where to start.
What is the development path for this specialist?

Answer 1

If indeed "In this I am a complete zero," then you have only one path. Hire a specialist (specialists!!!) to pay money and instruct them to solve your problem.
Because the path "I want to spend all my free time studying this subject area" is commendable. But dead end. After about 500 hours of study (in reality, I think that it is several times more, because judging by the style of the question, you are not only a complete zero in ML, but also in software development, systems analysis, and in a couple of other areas, which will have to be used - not for nothing I wrote "specialists" in the plural) - and so, after you find this amount of "free time" and spend it completely on obtaining the relevant knowledge, I think
And yes, your task has a very tangential relation to ML, in many ways other knowledge and approaches will really be required.
PS Well, why did everyone suddenly decide that studying ML on their own, in their free time and at home is easier than the same study, for example, geophysics or surgery ???