CleanedWeb

Use case

Job data from job boards and career pages

Collect job postings as records from an API: title, employer, location, posting date and the link to apply. Use them for a job board, a labor-market dataset or an alert for roles you track.

The problem

Postings are spread over many job boards and company career pages. Few of them offer a feed, and the ones that do rarely agree on field names. One board calls the employer "company", another "organization", and a third only shows it inside the title.

Aggregating them means a parser per site, plus a queue of fixes whenever a board changes its markup or pagination. Listings that expire have to be noticed and removed, or your data fills up with roles that no longer exist.

How CleanedWeb solves it

A website becomes an API. Point CleanedWeb at a job board or a careers page and it returns the postings as records with the same field names every time you call.

  1. Discover the fields

    CleanedWeb reads the listing and detail pages and proposes the fields it finds, such as title, employer, location, salary text and posting date. You decide which to keep.

  2. Build the API

    It builds and validates an API with inputs such as a keyword and location, and a limit on how many postings one call returns.

  3. Get a versioned contract

    The schema is saved with a version. If you aggregate several boards, you map each contract to your own posting model once and pin the version you mapped.

  4. Run it on your schedule

    Call the API from your own scheduler and store each response. Posting identifiers and URLs let you detect new roles and the ones that disappeared between runs.

  5. Repairs when the site changes

    When a board changes its pages, CleanedWeb rebuilds the API for it. A rebuild that would remove or retype a field is not swapped in silently, so your mapping does not change underneath you.

What a record looks like

An illustrative record from a fictional job board. Fields differ by site, and a field a posting does not show comes back as null.

{
  "id": "4821",
  "title": "Backend Engineer",
  "employer": "Example Robotics",
  "location": "Austin, TX",
  "remote": true,
  "employment_type": "full_time",
  "salary_text": null,
  "posted_at": "2026-01-12",
  "apply_url": "https://jobs.example.com/postings/4821"
}

Keep the posting URL or identifier as your dedup key, and keep the response from each run so you can tell a removed posting from one a run did not reach.

Docs for this use case

Free tools that help

Questions

Can I combine several job boards into one feed?

Yes. Each board becomes its own API with its own contract, and you map each one to a single posting model in your code. CleanedWeb does not merge boards for you.

Do I need the job board's permission?

CleanedWeb reads public pages. Whether you may collect and reuse a site's postings depends on its terms and on the law where you operate, so check both before you build. A site that requires sign-in cannot be turned into an API this way.

How do I find out a posting was removed?

A call returns the postings visible now. Compare them with the ones you stored from earlier runs; a posting missing from a complete run is a candidate for removal. Check each run's summary first, because a partial run is not a complete list.

Is there a ready-made API for a job board?

The API catalog lists published job APIs when there are any. If your board is not there, start a build for it.

Ready-made APIs in the catalog

These published APIs already return records for this kind of work. Open one to see its fields and a sample record.

Browse the full catalog

More use cases

All use cases

Turn a site into an API

CleanedWeb turns a website into a maintained REST API that returns structured records, with a versioned schema and repairs when the site changes.