Using AI to match candidates for job post

Need help
1

I run a marketplace, and I am looking for a solution to show customers relevant candidates based on their job postings. Some of the bubble marketplaces are already doing it, and I haven’t found something that addresses this specifically.

How can I achieve it in Bubble? If anyone is already doing it, please shed some light on me.

2

I’ll just describe it the other way round (users looking for jobs rather than jobs looking for applicants) as that’s what I’ve done before:

  • Get the description/skills/resume of applicant and get embeddings
  • Upsert to Pinecone using the unique ID of the applicant
  • Add a backend trigger such that when the applicant is modified the vector is deleted then reupserted/embedded
  • Get embeddings the job description + title
  • Query the Pinecone index using the embeddings of the job (applicants with profiles most similar to the job description)
  • Return the top few results (the ID of each result will be the unique ID in the database so Do a search for Applicants where unique ID = Pinecone query matches:each item’s id or whatever the exact expression is.
2 Likes
3

Sorry to response quite late, thanks for detail explanation
may i know what you were using for embeddings OPENAI ?
or is there any other which you can recommend ?

4

This topic was automatically closed after 70 days. New replies are no longer allowed.