I agree. Since you'll need to maintain the Python code for your data analysis, then why not Django for the back-end? It is a great framework comes with a lot of ready-to-use features.
If you are already committed to Python for the data analysis then it would make sense to use Django for the framework and stick with Python throughout. In general the fewer technologies you use (unless there is real justification) will improve process and long term costs.
All things being equal the fact you would not have less (likely no) licensing costs with a Python based solution will overall reduce your long term costs.