Employer ATS Job Feed — Free Beta

Collect job openings from companies that use Lever for their careers pages, and turn them into a tidy list you can download and review.

What is a Lever board? Lever is hiring software companies use to publish job openings. A Lever board is a company’s public jobs page hosted by Lever. This beta works with public Lever pages only, not every company’s careers website.

Get a small, filtered snapshot of published jobs directly from employers' Lever boards. Supply one to three employers and role filters; receive factual job metadata, public application links, observation timestamps, and a source-completeness report.

Use it to check a short recruiting or job-search watchlist and export matching roles to a spreadsheet. Employer names and board mappings come from your input. Verify the connection on the employer's own careers page before adding a board.

Open animated demo · The preview on this page is static; the link opens the looping GIF.

Fictional demo with an illustrative output layout. Public Lever boards only; employer mappings require your verification.

Try the prefilled example

The input form is prefilled with Metabase and engineering/account-executive title filters. Run it as supplied, then open the default dataset and the REPORT output. On October 7, 2026, the example returned eight matching jobs from 18 published postings through one public API request. Published jobs change, so future results can differ or be empty.

{
  "sources": [
    {
      "company": "Metabase",
      "domain": "metabase.com",
      "career_url": "https://jobs.lever.co/metabase",
      "identity_evidence_url": "https://www.metabase.com/jobs"
    }
  ],
  "filters": {
    "title_any": ["engineer", "account executive"],
    "title_exclude": ["intern"]
  }
}

For another employer, provide its hosted Lever career_url, or provide ats: "lever", its board identifier, and optionally region: "eu" (the default is global). Each source also requires company, domain, and identity_evidence_url. The evidence URL records your supplied mapping; this Actor does not independently verify it or discover boards from domains.

Filters and results

Supported filters are title_any, title_exclude, location_any, department_any, and workplace_any. Filters use case-insensitive literal substrings. Alternatives within each positive list are OR; different dimensions are AND. Any title_exclude match removes a role. An empty positive list does not restrict that dimension. To get all titles, set "filters": {}. Missing filters use the prefilled example filters.

The dataset contains matched rows with a stable job key, company and board, title, locations, department, workplace/employment metadata when supplied, public job/application URLs, observation dates, filter results, and source completeness. Repeated native IDs on the same board are deduplicated; similar jobs with different IDs can remain distinct.

REPORT records each source's dated observation, request receipts, completion status, errors or caps, and output counts. Inspect it alongside the dataset. A successful Actor execution can contain an incomplete source.

This beta is a fresh snapshot on every run. change: "baseline" means first observation in that run; it does not mean newly published. It does not compare earlier runs or report closed jobs. A fetch failure or capped result never establishes closure. Presence on a published board does not prove that an employer is actively reviewing applications. Native publication dates may be absent.

Scope and limits

This beta supports public Lever global/EU published-posting endpoints. Lever's official API documentation describes these endpoints and third-party access to published postings. Outputs contain factual metadata and links. Full descriptions, excerpts, raw responses, and a shared historical corpus are excluded. Applications stay on the employer's public application page.

Limit	Maximum
Explicit boards	3
Public API requests	3 pages per board; 9 total; no upstream retry
Jobs inspected	250 per board
Source response body	4 MB per board, plus a one-byte cap-detection sentinel
Output	750 rows and 1 MB of projected dataset JSON
Input	64 KB; 30 phrases per filter; 200 characters per phrase
Request timeout	15 seconds
Collection / process deadline	150 / 180 seconds
Memory setting	512 MB

Large boards can produce partial snapshots. Narrow the watchlist if output limits are exceeded. Unsupported providers and input options are rejected. Greenhouse, Ashby, general web crawling, login-only sources, and application submission are outside this beta's scope.

Free-beta pricing and feedback

The developer fee is $0. Apify platform usage consumes the runner's credits and can incur charges under their plan, including compute, storage, and transfer. Free beta does not mean free platform usage. Hosted cost and cold-start time still need measurement; the local example's timing is not a cloud performance promise.

For feedback, open an Actor issue with the public board URL, role filters, run ID, and expected result. Useful reports include incorrect mappings, missing fields, duplicate candidates, partial results, and time saved compared with a manual check. Do not include credentials, candidate data, or application details.