You might consider rephrasing your question, since individual raindrops are far smaller than the resolution of current sensors - so the best one could do is to infer the amount of rainfall from the signature of incidental phenomena. That involves considering the complete hydrological cycle, each phase, extant in whatever area of interest. This might yield some possible signatures, which would indicate which possible wavelengths available from a multitude of sensors on hundreds of orbital platforms - and then the observational characteristics of those platforms like coverage and repeat rate.
There are probably hundreds of knowledge centers dealing with this and related topics - the usual NASA and ESA portals ( https://science.nasa.gov/earth-science/focus-areas/water-and-energy-cycle ), but globally also international, national, state, and even very local metropolitan remote sensing data and analysis aggregators for very local conditions.
The use of Python or any other language is sort of irrelevant, because the data formats themselves are readable by readily available tool chains, and some data and analytical repositories use other languages, lie Google Earth Engine ( https://developers.google.com/earth-engine ) also use JavaScript to construct the queries and analytic chains.