how is the GenAI accuracy compared to AWS Textract?
Also I am not sure the libraries used by OpenAI get details from image input.
I think that understanding page layout of image is very important, which might not be the focus of OpenAI.
Looks awesome!
How is the response you get getting from the users?
Also for Hasura use case, does a BOT on docs sufficient. Asking because as per my understanding, a lot of query would be around troubleshooting.