Hi guys,
First of all, thank you in advance by your support. I’ve been reading about RASA and I’m working on a nice personal experiment, an autonomous robot which is not connected to internet (i.e. a Mars robot which could be operated using Natural Language commands such as "look for a black stone, move there and extract a sample). I have some professional background on space and I think it’s an interesting research topic since it could be useful for future space operations on asteroids and planet exploration.
I have some workarounds that I’d like to demo on Raspberry devices and I think RASA can be the right solution for it.
But the thing is once that I decided put my hands on the mug, I’ve started being stuck, specially with template actions. I’ll tell you what I’m looking for so you can tell me whether is crazy or simply not feasible, or maybe there is a better solution.
My first short term goal is having a raspberry device I can chat with, so after that, I can operate devices connected to such raspberry. I’ve created a chit-chat domain with their stories, templates and so on… things are running nicely and as expected when I execute rasa shell.
However, I’m trying to connect a basic front-end (using React) with this chit-chat through a call to the api (rasa run --enable-api --cors *). I can successfully parse the NLU part and I detect the intent, but I’m completely stuck trying to find the template message back. Just for the record, I’ve tried by installing the rasa-webchat package (https://github.com/mrbot-ai/rasa-webchat) but seems to be hanging eternally (I had it running an entire night because I read somewhere the installation took a long while).
I’ve been taking a look on the HQ/Rasa-Demo project but it didn’t offer much light on it (apart of there are nice ways to implement the calls on the Actions). Could you please light me a little? I would really appreciate.