This project demonstrates a number of concepts:
- NLP (or a simulation of it, using Eliza)
- NLP (or a simulation of it, using Roberta)
- Chatbot interactions (using REST API)
- Microservices frameworks (Helidon and Quarkus)
- And also using a simple
python
Flask app - Communication between three apps via a central app
- Flexibility to add more chatbots
- Ability to change the models used in place of the existing ones
git clone https://github.com/neomatrix369/chatbot-conversations.git
cd chatbot-conversations
./docker-runner.sh --runContainer
- Helidon world (Java)
- Quarkus world (Java)
- Roberta world (Python)
- Connecting worlds (Java)
Tweet: Making chatbots have a conversation video
Video: 'Chatbot Conversations' with Mani Sarkar and Yugo Sakamoto
See the different worlds (chat components) in conversation with each other:
See how you can make the different worlds chat to each other, go to How to build and run the projects.
You can also find out the underlying components that make up the NLP aspect of the chatbots there.
See how you can build and run your own docker container in your machine, go to How to build and run docker container.
Credits to the original authors of the code (different projects) used in this project. Special thanks to @yugoccp, @brjavaman and code4.life for giving us support for this community initiative!
This a demo created for illustration purposes, please don't expect a fully functional app or high quality code. The focus has been to use simple ready-to-use components and illustrate the original ideas via a new app.