How to build a chatbot on RASA

Question

I am building my own chatbot for culinary recipes. Right now I am using RASA as a framework, although I would like to learn more about the capabilities of Langchain or the ChatGPT API. My goal is to build new culinary recipes from other recipes taken in training, but I am not sure if this is the correct path. Does anyone have experience with this and can help me build such a project?

I am trying to build my chatbot on RASA and I expect this framework to help me design it in the best possible way.

Answer 1

My goal is to build new culinary recipes from other recipes taken in training

Depending on how you plan to implement this, Rasa may not be the best approach.

The training data in Rasa is used to boil down user input into intents and entities, and to provide specific conversation structure. The Rasa NLU Training Data Documentation provides detail on training data and other core components of the framework.

For dynamically creating new recipes based on other recipes in training data, ChatGPT or LangChain would be better options.