This video has been in the works for a while now, but I was finally able to piece together a FULL-BUILD tutorial that showcases how to integrate OpenAI into your app using as many OpenAI actions and columns as possible!
This demo app uses AI to generate ALL the content based on user input. I hope it’s instructive and approachable.
I like how you built the chat. I do it differently using two tables, a working table and a log. I’d be curious to see what your Chat History table looks like after a few questions were submitted.
Great stuff Bob. I am half way through the whole thing and learnt some tricks. Thanks for this…
Just a quick question…
Do you have an idea how much of the updates quota it uses up for that 1 recipe.
I was think that maybe since everything happens instantaneously does that mean it will use only 1 update or everytime a signal is sent to OpenAI, that uses the update quota?
Good question. After 3 hours of testing and tinkering during filming of this video, it used 47 updates…not too shabby for an extreme use case where EVERYTHING is generated using AI.
Thanks for the information, not bad at all but it would have to be a “paid for” App for it to work otherwise things can get messy especially if it’s a public facing App.
But I believe there is a business case to be made. We just need to think hard around the numbers and pass the costs over to the users in a way that it seems negligible to them and they are happy to pay for it because it solves a “pain point” for them.
Right. In my scenario, I might want to add a Wait action before the generate image action because it takes a bit to generate the title that the image is crafted from