Where Oink finds jobs—and what source coverage means

No single board shows the whole market. A big board has volume and crowds of applicants. A specialist board has fewer roles that fit better. A company's own careers system is usually the closest thing to the live source. A friend who wanted English-speaking, remote work found that those two conditions cut the number of jobs a lot. Oink searches across all these kinds of place and then ranks what it finds against your profile.

The source list changes as providers, parsers, and our own settings change, so this page doesn't give a fixed number. "Supported" also doesn't mean "in every run". It means Oink has a way to read that source when it is switched on, reachable, and relevant to your search.

Broad and startup job discovery

The current mix includes LinkedIn-derived public searches, Hacker News "Who Is Hiring?", Wellfound, Y Combinator startup jobs, Built In, and Welcome to the Jungle. They overlap, but not perfectly. A startup might post in a monthly thread, publish through its own ATS, and have a copy on a broad board, each at a different time.

That overlap is why deduplication matters. Oink aims to keep one useful record per job and keep the original posting URL and source name with it. A copied version can lag after the employer edits or closes the role, so trust the original page.

To search the monthly thread by hand, see how to search Hacker News "Who Is Hiring?". Developers who want structured public listings can read how LinkedIn collection works without a logged-in session, limits included.

Remote, regional, and specialist boards

Remote-focused discovery can include Remote OK, Remotive, We Work Remotely, and curated remote-board feeds. Regional technology sources include Just Join IT, No Fluff Jobs, InfoJobs, and Tecnoempleo. AI, academic, research, international-development, and public-interest coverage can include AI Jobs, EURAXESS, jobs.ac.uk, Academic Positions, Ikerbasque, EURES, UN Careers, ReliefWeb, Impactpool, Devex, and university-specific sources.

Each kind fixes a different gap. A remote board suits distributed work, but a listing can still carry country or timezone limits. A university page may publish a funded role that never reaches a commercial board. A regional site may show local language and work-authorization details that a bigger aggregator drops.

If location is your hard constraint, read how to check real remote-job eligibility. A remote label is a clue. It doesn't prove the employer can hire in your country.

Employer career systems and the open web

Many companies publish through applicant-tracking systems such as Greenhouse, Lever, Ashby, or Workable. Open-web discovery can find those employer pages and other public postings from role and keyword phrases built out of your profile. Oink does not send your raw resume to the web-search provider. The privacy policy explains that derived queries may be shared to run the search.

Employer pages sit closer to the company, and they still get things wrong. A requisition can stay public after an internal change, and a check that a job is still open can itself fail. Treat those checks as a way to lower risk, not as proof. Before you spend an evening on an application, open the original URL and confirm the role, the location, and the apply flow. The ghost-job checklist takes two minutes.

Why a named source might give you nothing

A source can return nothing useful for several ordinary reasons:

• your search is narrow and the source has no new matching roles;
• the source is rate-limited, blocked, or having a bad day;
• a quota, cooldown, or cost cap limits how much is fetched;
• the records duplicate a better copy from another source;
• the roles fail your location, seniority, or work-mode constraints;
• the source is on in general but isn't part of that search path.

Oink's adaptive source budgeting looks at how much a source supplies, how relevant it is, how often its jobs were delivered before, and its health. So the source mix moves, and no logo has to show up every day. Provider or model failures can also thin out the enrichment, so not every fetched record has the same depth of detail.

For developers and automation teams

If you want job data and not a personal shortlist, the operator also publishes nomad-agent Actors on Apify. Those are separate products with their own inputs, limits, run evidence, and billing. Some Oink articles link to them because they expose the collection layer directly, and the About and editorial standards page discloses that relationship.

Start with searching multiple job boards through one API call to see the bundle pattern. Check each Actor's current Store documentation before you integrate it, because marketplace descriptions and supported sources can change independently of Oink's consumer pipeline.

For job seekers, the simpler path is how Oink works: write down your constraints once, let the available sources feed one ranking, and check the evidence before you apply.