How to use spreadsheet with FAQs as database so that the bot autogenerate intents instead of writing stories

I’m builidng a faq bot using rasa. I have a excel sheet with questions and answers. I want to use that sheet as database so that bot makes it’s own intents and training data by using it. Is it possible. Please suggest any alternative.

EDIT: For better understanding Usually we manually write all the intents in nlu.yml file and identify intents and entities from a given sentence(s). But it is a huge task for large data set. So Is it possible somehow that I just give the paragraph and it automatically makes new intents which are possible from the para.

Yes, you can write a custom importer

Thank you. Will try that.

Can you please elaborate this? I have looked up custom importer but couldn’t find how it helps with my problem.

The docs show an example using a custom importer along with the standard Rasa file importer here.

In your case, you want the nlu data to come from your spreadsheet. To do that, you’ll need to add entries to Rasa domain and nlu configuration with your intents and example utterances.

The example python code has methods called get_domain and get_nlu_data. You’ll need to write your own version of these that reads your excel file and returns the intent info in yaml.