During my academic history with GIS/remote sensing/Geospatial (and still going on), I have seen how the same tools around the various fields in got plugged to i

During my academic history with GIS/remote sensing/Geospatial (and still going on), I have seen how the same tools around the various fields in got plugged to in, mostly for the data pre-processing, processing and analyst. The sacro-saint trilogy, for me at leat, is GDAL/OGR, PROJ and GEOS. They are the building blocks behind QGIS, PostGIS and most of spatial libraries in various programming languages (Rasterio and Fiona in Python for example). On a more GIS analyst perspective, most of it can be done with ESRI or QGIS and if you can get you hand dirty PostGIS is an amazing tool that I which I had the time to invest way more (or even Spatialite, sqlite flavour of GIS in database). I really which than Shapefiles goes away for good and be superseded by GeoPackage coupled with GeoJSON. The web had a hard time to be a proper tool for visualisation on GIS/remote sensing due to the big size of the various files but Cloud-optimized GeoTiff and Vector Tiles are the game changers but came too late. Commercial vendors tends to go for the cloud and platform lock down as Planet and their super-powered jupyter notebook and access to the catalog. Same for Google Eart Engine. We see also stuff like CartoDB and the Database As A Service for geodata. In a world when Geodata are highly available with better resolution than never, they are so big to store somewhere than it can be crazy. Working on time series at country scale (or even world scale) require now a lot of store space and calculation power for high resolution images (Even just Sentinel or Landsat at 10m and 30m resolution respectively). Small-scale is easier than ever and lower resolution can be managed on a laptop. I remember when I had to work with time series dataset of Sentinel-1 and -2 images when we were preprocessing 10-12 Tb of data with a pretty long chain of operations to do before handing it out to another team for furter data analysis. Some academic lab are storing Petabytes of remote sensing data gather during the time. "Geospatial Industry" is way more than looking at ESRI or Google Maps/Earth. A lot of things can be and are done with various research and industry projects than require high integration with the spatial field and the temporal analysis. And the quality of what you have as tools vary highly between field, actually I am working for a PhD on a Bayesian spatio-temporal modelling with lattice data. It is the wild west and you need to cure a lot of R libraries to get what you need. I don't know how much you can do in Python because we had to avoid any MCMC method due to constraints on processing power available : low in fund and no easy access to calculator, we settled for R-INLA who is very interesting on its own.