Yes, that is certainly possible. However, the csv files I use are 30GB and this is also the space that is needed in Azure SQL. Without optimization (e.g. omit large description column I use), this would mean that you also need a significant cache size in Azure Redis and that is costly. If you take the standard SKU of 56GB, that would be 1500 euros per month, see https://azure.microsoft.com/en-us/pricing/details/cache/. In comparison, the Azure SQL 125 DTU standard SKU I used, is 150 euros per month.