I found the distribution of my data with "distfit" library for python. But what now? The best distribution that describes my data is "weibull" distribution. But I don't know what can I do with this knowledge. Can someone help?
Imho there's probably nothing to do with this information, especially considering only the technical side of it.
It's highly subjective, but I think that what people mean when they say to "know the distribution of your data" is that it is useful to have an intuitive understanding of what your data consists of: main stats, characteristics, how much variance, imbalance, important patterns between variables, etc. This information, put together with the expert knowledge related to the specific task at hand, would normally help an experienced data scientist decide the design of the system (what kind of algorithm, preprocessing, etc.).
But it's not a recipe, you can't expect to follow deterministic steps like with a manual. It's more an analysis depending on the context, the time one wants to spend, etc. My advice for improving the performance of any system: start by investigating a sample of the errors it makes. See if these errors are preventable (they might not be), and if yes what prevents the system to find the correct answer.