I’m a beginner with no code and I’m trying to fix api calls for AI but I got some problems with that.
I’m tryin to create an AI integrated app for CVs assestment.
So, the AI should take informations both from CV (pdf file) and job description of that particular job.
The results I want back are:
1- Is there a personal photo in the pdf? yes or no
2- a number from 1 to ten that tells me how the candidate fits with job position
3- A brief description about the candidate.
I’d like these results to be shown in my candidate tab.
Any suggestions to fix my api call for my objective?
Thank you all
Sergio
P.S.
Se esiste qualcuno che usa Bubble e parla in italiano per favore palesati!
This sounds like a great idea and actually something related to what I’m working on.
For starters, have you decided which AI api you will use? I’ve used chatGPT with the API connector plugin and it works really well.
Just be aware that during your testing you’ll still be charged for api calls so I recommend setting it to one of the cheaper models like 4o-mini.
The PDF data extraction is actually the part which I’m not sure how to do, maybe there’s a way of getting the PDF source code and then passing that to chatGPT with a prompt asking ‘is there an image included?, give a rating 1-10 of how well it matches this job description.
I’m undecided between chatGPT and Claude, but I think I’m gonna choose the first one.
I think we got same issues:
How can I include booth pdf file and job description in api call?
How can I ask chatGPT what I want him to do with these informations?
How can I have the result where I want it in my Bubble app?
Do I have to convert to json before chatGTP api call?
All tutorials I found are about simple generation apps where input and output settings are unserstandable.
But nothing about what I’m tryin to fix.
What’s your experience about that?
Where are u from?
Why don’t either of you use Groq and a Llama model. It’s cheaper and faster than both.
You don’t even need to use an API. You can use the AI Bot plugin on the plugin store which will do 99% of the work your looking to achieve Inc rag and pinecone for the uploaded resumes.
Ho lavorato tanto nel Recruitment e ho integrato API per fare queste cose.
Et parlo un può italiano. Sono on Libkedin se vuoi parlare de questo Matthieu BOUILDÉ
