This is a demo weather assistant created with Rasa framework.

This system uses the OpenWeatherMap API to fetch real time weather data. You just need to sign up there and provide actions/actions.py with your personal API token.

Install Rasa in a clean virtual environment using pip.

$ python3 -m venv ./venv
$ source ./venv/bin/activate
$ pip3 install -U pip
$ pip3 install rasa

Train your model using:

$ rasa train

Start the API action server with which the model communicates:

$ rasa run actions

Talk with your bot throught command line with:

$ rasa shell

In the next graph all stories are visualized. The intents are shown in blue boxes and the bot's responses in white. The 2 stories on the left, starting with the greet intent are the default simple Rasa stories. Our weather assistant, the two created stories, is visualized in the right two pipelines.

This graph has been created with Rasa's internal command:

$ rasa visualize