About SWE Job Lists

How it works

1. 1

   Collect

   Every 2 hours we read the community-maintained GitHub lists: SimplifyJobs, jobright and speedyapply. Listings older than 30 days are dropped.

2. 2

   Fetch descriptions

   For each listing we fetch the posting from the employer's applicant tracking system and pull out its requirements. When a page can't be fetched, the job is handled from its title, company and location, and we say so wherever that matters.

3. 3

   Score with a calibrated decision model

   Each posting gets a public H-1B sponsorship estimate. If you sign in and add your resume, listings get a private fit score from 1 to 100: five dimensions (required skills 35%, experience level 20%, role similarity 20%, education 15%, domain 10%) rated by a decision model that answers with probabilities, not chat text, read in both orders and averaged, then capped at 25 if an eligibility check fails. Full details in the FAQ.