The best is to use whatever works for the use case. I prefer to use as less 3-rd party tools as possible. But better focus on getting data directly from DB in Python.
Such requirement is out of scope in this example — all TF’s are independent. However, thinking about your case — I think this can be achieved by calling JS from ADF Task Flow running in inline frame.
You can achieve it only by custom code implementation. You would need to handle failed classification requests, collect data and then call model re-train. In my new article, I show how to implement chatbot backend in Node.js, you can do all such custom logic there (in my example, I handle chat context there) —…
I wasnt using Oracle DB Python interface with TensorFlow. Technically it should work to install it using PIP from command line inside container: https://cx-oracle.readthedocs.io/en/latest/installation.html#install-using-pip