This repo contains blueprints for various conversational applications. A blueprint is a core reference implementation of a MindMeld conversational application. Blueprints can be leveraged to kick start development on a new MindMeld application with a similar use case.
Assuming you have pip installed with Python 3.4, Python 3.5 or Python 3.6 and Elasticsearch running in the background:
pip install mindmeld
python -m home_assistant build
python -m home_assistant converse
The Food Ordering blueprint is a conversational app for ordering food from a food delivery service such as Amazon Restaurants, Caviar or JustEat. This blueprint will demonstrate building an application for interacting with a number of restaurants with different menus.
In order to convert the food ordering application into a Webex Teams bot, first make sure you have ngrok (https://ngrok.com/) installed and do the following:
# Expose your local server's port 8080 to the internet
./ngrok http 8080
# Create web hook here: https://developer.webex.com/docs/api/v1/webhooks/create-a-webhook.
# Use ngork's publically accessible URL for the targetUrl param when creating the web hook and
# set resource=messages, event=created
export WEBHOOK_ID=<insert webhook id>
# Create bot access token here: https://developer.webex.com/my-apps/new
export BOT_ACCESS_TOKEN=<insert bot access token>
# Start the bot service
python webex_bot_server.py
The Home Assistant blueprint is a conversational app that helps users to control appliances, lights, doors, thermostat and check the weather. This blueprint will demonstrate building an application with multiple intents, domains, entities and roles.
The Video Discovery blueprint is a conversational app that helps users to search for and ask questions about movies and tv shows. This blueprint will demonstrate the extensive usage of the Knowledge Base component.
The HR Assistant blueprint is a conversational app that helps users to search for and ask questions about employees working in an organization. This blueprint will demonstrate building an enterprise facing application with multiple intents, domains, and entities. It will also showcase extensive usage of the Knowledge Base and Question Answerer Components.
Visit the MindMeld website.
Please contact us at [email protected].