Job Scraper That Never Gets Blocked — straight from the company's ATS

Pull every open role from any company's Greenhouse, Lever, Ashby, SmartRecruiters, BambooHR or Recruitee board, in one unified format. No login. No cookies. No browser automation. Nothing that can get an account blocked, because there is no account.

Optional change tracking tells you which roles appeared, closed or were edited since your last run — the difference between a job list and a hiring signal.

Why not just scrape LinkedIn or Indeed?

Because those break, and they break in the two ways that cost you real time:

They get blocked, and they put your account at risk. Anything scraping a logged-in surface is fighting anti-bot systems that change weekly. This reads the same public JSON endpoints companies publish so applicants can find their jobs. There is nothing to detect and nothing to ban.

Their output moves under you. The most-used LinkedIn job scraper on this store is rated 3.9, and its top review says it plainly: one week it strictly matches job titles and returns 250 postings, the next it returns 5,000 irrelevant ones. That makes it unusable in a pipeline — every run becomes a cleaning job.

This actor takes the opposite position, and puts it in writing below.

The schema contract

Every row carries schemaVersion. While that number stays the same:

• fields are added, never renamed, retyped or removed
• filters mean the same thing every run — titleFilter is a case-insensitive regular expression applied to the job title, and nothing else
• a field that cannot be determined returns null, never a guess

If a breaking change is ever genuinely necessary, the version increments and the change is documented. Your pipeline gets to keep working.

The problem this solves

Most companies publish open roles through Greenhouse, Lever, Ashby, SmartRecruiters, BambooHR or Recruitee, and each publishes a completely different shape. Greenhouse calls a link absolute_url, Ashby calls it jobUrl, Lever calls it hostedUrl. Dates arrive as ISO strings in some and epoch milliseconds in another. Departments are nested arrays in one, flat strings in the next. SmartRecruiters paginates and keeps descriptions on a separate call entirely.

Every other job scraper handles exactly one provider and hands you its raw shape. Track companies across all six — and you will, because companies don't coordinate their ATS choices — and you're left writing and maintaining that join yourself.

This does it for you. One schema, one table, every provider.

What you get

Field	Notes
schemaVersion	The output contract this row was produced under
company, companySlug, ats	Who, and which system published it
title, department, team	Normalised across all six providers
location, isRemote	See the note on remote detection below
employmentType	Full-time, contract and so on, where published
postedAt, updatedAt	Always ISO 8601, including Lever's epoch timestamps
jobUrl, applyUrl	Public posting and direct application links
description	Full text, HTML stripped
compensation	Salary range where the company discloses one

Honest remote detection

isRemote returns true, false, or null when it genuinely cannot be determined.

Ashby, SmartRecruiters and Recruitee publish real flags. Greenhouse and Lever only hint at it in free text, so a location reading "London" tells you nothing about whether the role is remote. Most scrapers return false in that case, which is a guess dressed as data — and it quietly breaks anyone filtering on it.

Null means unknown. Filtering with remoteOnly excludes unknowns rather than assuming.

Change tracking

Turn on trackChanges and schedule the actor daily. Each run compares against the previous snapshot and flags every role as appeared, closed or edited.

That turns a static list into a feed:

• A company opening four sales roles in a week is expanding a team
• A company quietly closing its entire platform team is a different signal entirely
• Roles that reappear after closing usually mean a failed hire

Nobody else on the store sells this, and it is the reason to run this on a schedule rather than once.

Input

{
  "companies": ["greenhouse:stripe", "lever:plaid", "ashby:ramp", "smartrecruiters:Visa", "bamboohr:jobandtalent"],
  "trackChanges": true,
  "remoteOnly": false
}

Finding a company's slug — it's the last part of their careers URL:

ATS	Careers URL	Slug
Greenhouse	boards.greenhouse.io/stripe	stripe
Lever	jobs.lever.co/plaid	plaid
Ashby	jobs.ashbyhq.com/ramp	ramp
SmartRecruiters	jobs.smartrecruiters.com/Visa	Visa (case-sensitive)
BambooHR	jobandtalent.bamboohr.com/careers	jobandtalent
Recruitee	personio.recruitee.com	personio

A bare slug like "stripe" defaults to Greenhouse. If you get the provider wrong, the error tells you exactly which URL to check rather than just returning nothing.

Filters — remoteOnly, titleFilter (accepts a regular expression, e.g. engineer|developer), maxJobsPerCompany, and includeDescription to shrink the payload when you only need titles and links.

Common use cases

• Recruiting and talent mapping — monitor hiring across a target list of companies
• Sales prospecting — a company hiring ten engineers is buying developer tools
• Competitive intelligence — track headcount direction by team
• Job aggregators — one integration instead of three
• Investment research — hiring velocity as a growth signal

Reliability

One company failing never fails the run. If you're tracking two hundred companies and one board is unreachable, you get one hundred and ninety-nine, plus a machine-readable record of which one failed and why. You're only charged for roles actually returned.

These are the providers' own public JSON endpoints — no login, no session, no browser automation — which makes this one of the most stable data sources available. Wrong slugs produce a clear error naming the URL to check, rather than a silent empty result.

That last point matters more than it sounds. SmartRecruiters answers an unknown company identifier with HTTP 200 and zero results rather than a 404, so a typo looks exactly like a company that isn't hiring. This actor tells you which of the two it is.

FAQ

Which job boards does it read? Company career sites running Greenhouse, Lever, Ashby, SmartRecruiters, BambooHR or Recruitee — the ATS platforms behind most tech and scale-up hiring. You name companies, it finds their jobs at the source.

Why is this better than scraping LinkedIn or Indeed? No login, no anti-bot wall, no account to get blocked, and no selector to break when a page redesigns. ATS APIs are public JSON the companies themselves publish.

Will the output format change and break my pipeline? Every row carries schemaVersion. Fields are added, never renamed or removed, without that number changing. That is the contract.

Can I get only remote jobs? Yes — and remote detection is honest: a job is flagged remote from the ATS's own fields, not guessed from the title.

Can I export to CSV or feed an ATS/CRM? Yes — CSV, Excel, JSON or JSONL per run, or pull the dataset via the API on a schedule.

How fresh are the results? As fresh as the company's own careers page — the ATS is where postings appear first, usually before the aggregators.

Notes

Data comes from publicly published job boards that companies operate to attract applicants. This actor is unofficial and is not affiliated with Greenhouse, Lever, Ashby, SmartRecruiters, BambooHR or Recruitee. Job descriptions remain the property of the publishing company — use them in line with applicable laws and terms.