Retail Example Bot

This is a sample retail bot to help provide starter content, examples of how to implement particular features, and sample use cases. Built using Rasa 2.3.1

Install dependencies

Run:

pip install -r requirements.txt

Run the bot

Use rasa train to train a model.

Then, to run, first set up your action server in one terminal window:

rasa run actions

In another window, run the duckling server (for entity extraction):

docker run -p 8000:8000 rasa/duckling

Then talk to your bot by running:

rasa shell --debug

Note that --debug mode will produce a lot of output meant to help you understand how the bot is working under the hood. To simply talk to the bot, you can remove this flag.

Overview of the files

data/stories.yml - contains stories

data/nlu.yml - contains NLU training data

data/rules.yml - contains the rules upon which the bot responds to queries

actions/actions.py - contains custom action/api code

domain.yml - the domain file, including bot response templates

config.yml - training configurations for the NLU pipeline and policy ensemble

tests/test_stories.yml - end-to-end test stories

Things you can ask the bot

Check the status of an order
Return an item
Cancel an item
Search a product inventory for shoes
Subscribe to product updates

The bot can handle switching forms and cancelling a form, but not resuming a form after switching yet.

The main flows have the bot retrieving or changing information in a SQLite database (the file example.db). You can use initialize.db to change the data that exists in this file.

For the purposes of illustration, the bot has orders for the following email addresses:

example@rasa.com
me@rasa.com
me@gmail.com

And these are the shoes that should show as in stock (size, color):

inventory = [(7, 'blue'),
             (8, 'blue'),
             (9, 'blue'),
             (10, 'blue'),
             (11, 'blue'),
             (12, 'blue'),
             (7, 'black'),
             (8, 'black'),
             (9, 'black'),
             (10, 'black')
            ]

Testing the bot

You can test the bot on test conversations by running rasa test. This will run end-to-end testing on the conversations in tests/test_stories.yml.

Note that if duckling isn't running when you do this, you'll see some failures.

Rasa X Deployment

To deploy this bot, it is highly recommended to make use of the one line deploy script for Rasa X. As part of the deployment, you'll need to set up git integration to pull in your data and configurations, and build or pull an action server image.

Action Server Image

You will need to have docker installed in order to build the action server image. If you haven't made any changes to the action code, you can also use the public image on Dockerhub instead of building it yourself.

It is recommended to use an automated CI/CD process to keep your action server up to date in a production environment.

srikanthrc / retail-demo