I’m trying to make an image recognizer in coda.
It would work like this:
User uploads a photo of a banana. AI in coda compares image to other images or bananas and says if image is a banana.
Is any of that possible? (I have the feeling that the chance of it not being possible is 99.9%).
Hey! I’ve actually managed to achieve something like this recently using Integromat and Google Cloud Vision. This is what the flow looks like in Integromat. When a new image is added to a particular folder in Google Drive, a new row is created in Coda. The image then gets passed through Google Cloud Vision and tags are created. These tags are then upserted into a master tags table in Coda. Once the master table has been updated, then the group of tags for that image are put into the row that was created when the image was uploaded.
Besides Coda, I think Integromat is my other favorite thing lol. Nice work with this! 
Matt- that is so cool.
WOW @mattbogaards! 
One thing though, can you list what each action does. e.g. For the text parser action, I don’t know what text it should parse.
