npm install in the root folder

install all the python dependencies,

run npm start with admin permissions

run python3 server.py

The ML model is running locally and the results are served on the path 127.0.0.1:5000/predict. POST request and takes JSON payload with input field which is the list of all input parameters
The PCAP packet tracer is defaulted to port 3000, use serve command tool to host any file in LAN and make requests to the specified serve URL to log packets and get predictions

Only works in a Linux environment as the PCAP needs Admin/root permits. WSL can be a great alternative