Just want to say a quick thank you for this tutorial. I just finally got around to implementing it in my app and it turned out to be amazing! Thank you sir!
20:29 Excited to watch this tutorial you mentioned just because in the example I'm trying to work on, there are like 40+ numbers/words I'd be grabbing!
Simply brilliant 👏 I have a question, what would be the advantage of using this instead of the new Assistants? Are assistants less trainable and less accurate? Because i know they only get information from file retrieval and i can't have it to output in a specific way or instruct it something special so i was wondering if custom gpts are better? Or is there a way to train the assistants beside file upload that im missing?!
Great tutorial thank's for sharing it ! I have one question, i'm trying to do it on my app who contain employee of the user do you think it could be possible to create a link between user and the shift using the Json format, to be more clear this is my use case. The user take a picture of a planning i would like to put it in Json and each "Shift" from this planning have a link to a user, so i'm thinking by looking by each first name and last name (because i don't think they would put the uid in the planning ahah) and connect the shift to the right user. Thank's again and all the best from France ! :)
Thank you very much for this video. If the app doesn't generate ALL the images, just a few every now and then, what can be the problem and how can I fix it? I made the index*30 to try and increase the time, yet it's still not creating all the images. Any help will be much appreciated.
Can get everything working except images unfortunately. Made sure the url type was image, made sure the JSON keys were consistent across the workflows. Not sure where it could be going wrong as I followed everything exactly. Any ideas?
Great tutorial, much needed. I am having an issue when shifting the OpenAI call to the backend workflows instead of running it client side. It works well client side, or when initializing the call, but I am getting an HTTP 400 error when doing it via a backend workflow. Any idea why this might be happening?
Managed to figure it out myself. For anyone else having a similar issue, the problem was with the syntax in the OpenAI API call. I had it as "", whereas it should be without the quotation marks as the format as JSON Safe adds the quotation marks itself. Completely overlooked it in the video.
@@admirleka7467 - thanks for posting the update. I was going to say - there's no reason why it should work on the client side and not work in your backend workflows (unless you're on a free plan in which case backend workflows don't work), so it's likely to be a small syntax error which I always get loads of. The amount of times I've forgotten to remove or add some quote marks!
If the plugin works and is maintained then not really, but the risk is that OpenAI makes a change to their API or adds new features which aren't supported by the plugin. On the whole if you pick a popular plugin you're probably fine.
An incredibly dense and rewarding video. As someone who just started Bubble, this was mind-blowing when I actually got it to work. Thank you!
Just want to say a quick thank you for this tutorial. I just finally got around to implementing it in my app and it turned out to be amazing! Thank you sir!
Can't believe there's so much value in this channel, amazing work, and very helpful!
20:29 Excited to watch this tutorial you mentioned just because in the example I'm trying to work on, there are like 40+ numbers/words I'd be grabbing!
That's great, it should be helpful, especially if you can loop through them.
Simply brilliant 👏
I have a question, what would be the advantage of using this instead of the new Assistants?
Are assistants less trainable and less accurate?
Because i know they only get information from file retrieval and i can't have it to output in a specific way or instruct it something special so i was wondering if custom gpts are better?
Or is there a way to train the assistants beside file upload that im missing?!
Amazing mate! really nice project!
Outstanding video thank you for sharing your knowledge !
thanks a lot for this tutorial, dude! Awesome!
This was amazing! Thank you so much for that!
That was awesome ! Thank you ! One question though : how do you keep the formatting from GPT4 when displaying the data back in your bubble app ?
Great tutorial thank's for sharing it !
I have one question, i'm trying to do it on my app who contain employee of the user do you think it could be possible to create a link between user and the shift using the Json format, to be more clear this is my use case. The user take a picture of a planning i would like to put it in Json and each "Shift" from this planning have a link to a user, so i'm thinking by looking by each first name and last name (because i don't think they would put the uid in the planning ahah) and connect the shift to the right user.
Thank's again and all the best from France ! :)
This is great. I am wondering if we can embed a custom gpt like botpress or voiceflow in bubble as the main feature? Thanks
Thank you very much for this video. If the app doesn't generate ALL the images, just a few every now and then, what can be the problem and how can I fix it? I made the index*30 to try and increase the time, yet it's still not creating all the images. Any help will be much appreciated.
Brilliant tut… advanced but well executed bravo! Thanks much
You're very welcome!
@@nocodelife Could you please share a link to your editor for this? I would truly appreciate it. Thanks!
What’s the issue you’re having?
@@nocodelife No issues at the moment, just for reference when I start integrating it in my own app if ok.
Excellent!!!
Would love to see a full AI course.
Thanks, what kind of things would you like covered in it?
@@nocodelife Something around leveraging Custom GPTs from the OpenAI API :)
@@nocodelife Connecting telephony and the ability to chat, or continue communicating with artificial intelligence, by phone
Can get everything working except images unfortunately. Made sure the url type was image, made sure the JSON keys were consistent across the workflows. Not sure where it could be going wrong as I followed everything exactly. Any ideas?
Fantastic!
Many thanks!
Great tutorial, much needed. I am having an issue when shifting the OpenAI call to the backend workflows instead of running it client side. It works well client side, or when initializing the call, but I am getting an HTTP 400 error when doing it via a backend workflow. Any idea why this might be happening?
Managed to figure it out myself. For anyone else having a similar issue, the problem was with the syntax in the OpenAI API call. I had it as "", whereas it should be without the quotation marks as the format as JSON Safe adds the quotation marks itself.
Completely overlooked it in the video.
@@admirleka7467 - thanks for posting the update. I was going to say - there's no reason why it should work on the client side and not work in your backend workflows (unless you're on a free plan in which case backend workflows don't work), so it's likely to be a small syntax error which I always get loads of. The amount of times I've forgotten to remove or add some quote marks!
Is there any benefit or drawback of using the API connector vs using one of the openAI plugins?
If the plugin works and is maintained then not really, but the risk is that OpenAI makes a change to their API or adds new features which aren't supported by the plugin. On the whole if you pick a popular plugin you're probably fine.
@@nocodelifeCan you inspect all of this plugin’s source code? How can we trust them to not have potential security vulnerabilities? Thanks!