sahilsk11
commented
3 years ago Hi William,
Thanks for the kind words! You're absolutely correct that this tool is built on top of the Lever API, to which we no longer have access. We emailed Lever's CEO, but he shut us down, saying this API is only available for internal use. Apparently this API was never supposed to be public - it was chance we found it in another project and were able to use it.
@Saaniya00 is currently spearheading the search for an alternative. Ideally, we'd like to use an API provided by popular ATS's like Greenhouse or Workday to ensure that we use the most accurate parsing systems - ones that candidates will encounter when they apply. We're still looking for good alternatives on this front.
If you (or anyone reading this), has suggestions for API's that work well, we'd love help on this front. We're looking for a public API that can return a JSON-parsed version of a resume file, with previous roles, schools, etc. cleanly parsed from a resume.
This has the potential to be a game changing project for Computer Science students. I understand how important this could be as someone who has spent hours optimizing resumes, writing cover letters, and applying to every internship I could.
After cloning the repo myself and having a look I quickly realized what you have built is really just a beautiful front-end layered on top of multiple third party APIs. Your resume_to_dict function is making a request to https://jobs.lever.co/parseResume.
You are obviously aware this returns an authentication error. This is certainly not a call to their official API so I am assuming you discovered you could piggy back off this exposed endpoint.
Are you planning on rectifying this issue by moving the work this endpoint originally completed to your own backend? Is it feasible to explore a different option? @sahilsk11 , you are obviously very talented and motivated! I know with a little work you could tackle the challenge.