This is a really interesting project! From what I can tell the current training data is AWS and GCP focused, have you thought about opening that up to CNCF white papers? I was recently involved in the platform white paper, but there are a number of really good ones around operators, gitops, security and more.
Just to say, I realise that โwhite paperโ can often be a code word for marketing, so I like that it is limited to trusted sources, but I am just proposing that the CNCF may be an equally trusted source to AWS and GCP.