Career Site Jobs Scraper — Workday, Greenhouse, Lever & 6 more

Every open job of the companies you list, read live from their own job boards: Greenhouse, Lever, Ashby, Workday, SmartRecruiters, Recruitee, Workable, Personio and Rippling. Give a link, a website or just a name. All nine come out in one row shape, with salary as numbers, and each company ends with a plain check: how many jobs its board says it has, and how many you received.

Versión en español más abajo.

Quick facts

• What it does: Every open job of the companies you list, read live from their own job boards on 9 ATS (Greenhouse, Lever, Ashby, Workday, SmartRecruiters, Recruitee, Workable, Personio, Rippling), in one row shape, salary as numbers, Workday past its 2,000-job stop.
• Use it for: tracking the jobs of a list of target companies; hiring signals; job feeds without a database lag; monitoring only new jobs.
• Not for: job boards that aggregate many employers (Indeed, LinkedIn); BambooHR, Teamtailor, Breezy, iCIMS, Oracle or SuccessFactors boards (not in this version); finding a Workday site by name alone.
• Input: companies (required): job-board or careers links, ats:slug (e.g. greenhouse:stripe), a website or a plain name; Workday needs its link. Optional filters (applied before charging): titleKeywords, excludeTitleKeywords, locations, workplaceTypes, departments, postedWithinDays; maxJobsPerCompany (0 = all); onlyNewJobs + stateName; includeDescription (default on).
• Output: one dataset row per job, and one SUMMARY record with "ATS says N, delivered N" per company. Every field with its type is under Output reference below.
• Charged events (pay per event): job USD 0.0015 per job delivered (USD 1.50 per 1,000), full description included; plus Apify's standard start event apify-actor-start.
• Never charged: companies as such (there is no fee per company), a company whose board is not found, jobs your filters drop, jobs whose page could not be read, jobs already delivered in monitoring mode, the SUMMARY record.
• Limits: stops at maxJobsPerCompany or at your maximum charge per run (SUMMARY stoppedBy); a board found by name says so (detectedBy): check the company column; do not run two monitors with the same stateName at once.
• Run it from: Apify Console, the Apify API or an AI agent through the Apify MCP server (Actor pampadata/career-site-jobs-scraper), with schedules and integrations.

Minimal input

A small input that works, taken from this Actor's own tests; every other field is optional:

{
  "companies": [
    "greenhouse:datadog"
  ],
  "titleKeywords": [
    "engineer"
  ],
  "excludeTitleKeywords": [
    "manager"
  ],
  "locations": [
    "New York"
  ],
  "postedWithinDays": 60
}

Why this one

• Live, not a database. Each run reads the company's board at that moment: a job posted ten minutes ago is there, a closed one is gone. No 6-month window, no "company not in our index".
• Workday past its 2,000-job stop. Workday never lists more than 2,000 jobs per search (later pages repeat the first ones). This Actor splits the search by Workday's own facets (job family, time type, location...) and merges the parts. Tests on 2026-10-08:

Workday site	Workday's own count	One plain search	This Actor
NVIDIA	2,000 (facets: 2,680)	2,000	2,680
PwC (experienced)	3,855	2,000	3,855
Target	2,000 (facets: 14,149)	2,000	14,130

Another Workday Actor in the Store, run on NVIDIA the same day, returned 2,000 jobs and reported 2,000 jobs on the site.

• The check per company. The SUMMARY record says, for every line you gave: which board it matched and how, how many jobs the board says, how many were listed, filtered out and delivered, and why a company is not complete if it is not.
• Salary as numbers (minimum, maximum, currency, period) from the ATS's own salary field where it has one (Ashby, Lever, Recruitee, Rippling, SmartRecruiters) and otherwise from the pay-range sentence in the ad (US/EU pay-transparency text: "Pay Range $102,000 — $120,000 USD", "148,000 USD - 235,750 USD"). In tests: Datadog 217 of 432 jobs with a salary, OpenAI 651 of 812 (from Ashby's field), NVIDIA about half (from the text).
• Only new jobs for monitoring: run it daily on the same list and you get (and pay for) only the jobs you have not received before, plus a count of jobs closed since the last run.

What you can give it (one per line)

You type	What happens
https://job-boards.greenhouse.io/stripe, jobs.lever.co/spotify, jobs.ashbyhq.com/openai, https://nvidia.wd5.myworkdayjobs.com/NVIDIAExternalCareerSite, jobs.smartrecruiters.com/ServiceNow, bunq.recruitee.com, apply.workable.com/huggingface, acme.jobs.personio.de, ats.rippling.com/rippling/jobs	That board, directly. A link to one job or to the API works too.
greenhouse:stripe, lever-eu:mistral	That board, directly.
figma.com	The Actor looks for a job-board link on the site (home, /careers, /jobs); if there is none, it tries the domain's name on each ATS.
Datadog	The name is tried as a board name on the 8 ATS that use one (Workday needs its link). The board with the most jobs wins; the others are listed in SUMMARY as alsoFoundOn.

A board found by name says so in SUMMARY (detectedBy), with a warning to check the company column: two companies can share a name. In a test of 20 well-known companies, 17 were found by name alone; the other 3 use Workday or their own system and need a link.

Filters (you pay only for jobs that pass)

Title contains any of / does not contain (whole words: "data" matches Data Engineer, not Database), location contains (or country code: US, DE), workplace type (remote, hybrid, on-site), department or team, posted in the last N days, maximum jobs per company.

What you get for each job

company, ats, board, jobId, title, url, applyUrl, department, team, location, locations (all of them), countryCode, workplaceType (REMOTE / HYBRID / ON_SITE), employmentType (FULL_TIME, PART_TIME, CONTRACT, TEMPORARY, INTERN) and the ATS's own words, postedAt, updatedAt, salaryMin, salaryMax, salaryCurrency, salaryPeriod, salaryText, salarySource (ats or description), requisitionId, descriptionText, descriptionHtml, contactsRemoved, input (your line). Fields an ATS does not publish stay empty; nothing is guessed.

Example (trimmed)

{
  "company": "openai",
  "ats": "ashby",
  "board": "openai",
  "jobId": "fc5bbc77-a30c-4e7a-9acc-8a2e748545b4",
  "title": "Data Engineer",
  "url": "https://jobs.ashbyhq.com/openai/fc5bbc77-a30c-4e7a-9acc-8a2e748545b4",
  "applyUrl": "https://jobs.ashbyhq.com/openai/fc5bbc77-a30c-4e7a-9acc-8a2e748545b4/application",
  "department": "Applied AI",
  "team": "Applied AI Engineering",
  "location": "San Francisco",
  "locations": [
    "San Francisco",
    "Mountain View"
  ],
  "countryCode": "US",
  "workplaceType": "HYBRID",
  "employmentType": "FULL_TIME",
  "employmentTypeRaw": "FullTime",
  "postedAt": "2023-09-26T22:34:48Z",
  "updatedAt": null,
  "salaryMin": 235000,
  "salaryMax": 385000,
  "salaryCurrency": "USD",
  "salaryPeriod": "YEAR",
  "salaryText": "$235K – $385K",
  "salarySource": "ats",
  "requisitionId": null,
  "descriptionText": "ABOUT THE TEAM\n\nThe Applied team works across research, engineering, product, and design to bring OpenAI’s technology to consumers and businesses.\n\nWe seek to l…",
  "contactsRemoved": 0,
  "input": "openai",
  "descriptionHtml": "<p>…</p>"
}

The SUMMARY record

In the run's key-value store, one entry per line you gave: ats, board, boardUrl, detectedBy, alsoFoundOn, atsSays, listed, filteredOut, alreadyDelivered (only-new mode), closedSinceLastRun, delivered, complete, stoppedBy, warnings, errors. It is saved as the run goes, so a run stopped by its timeout still says what it got. The same board given twice (a name and a link) is delivered and billed once.

Only new jobs (monitoring)

Turn on Only new jobs and give a State name (one per monitor). The first run delivers every matching job; later runs with the same state name deliver only jobs not delivered before. The memory is a named key-value store, career-site-jobs-state, in your own Apify account (delete it to start over, or use another state name). A job that was filtered out or cut by your limits is not remembered, so it comes when it matches. Each delivered batch is remembered at once, so a run that stops halfway does not bill those jobs again. Do not run two monitors with the same state name at the same time: both would see the same jobs as new.

Pricing

You pay only for jobs delivered: USD 1.50 per 1,000 jobs (USD 0.0015 each), full description included. Nothing per company, nothing for a company whose board is not found, nothing for jobs your filters drop, nothing for a job whose page could not be read. Apify also bills its standard tiny start event (USD 0.00005 per run, per GB of memory).

The Actor stops cleanly at your maximum cost per run, and the SUMMARY record says where it stopped.

Good to know

• Workday needs the career-site link (https://<company>.wd<N>.myworkdayjobs.com/<site>): the site name cannot be guessed. Open the company's careers page, click a job, and copy the address up to the site name.
• Workday, SmartRecruiters and Rippling publish the full text, exact date and salary only on each job's own page, so the Actor opens it (one request per job). Turning off Include the full description skips those requests (much faster on big Workday boards; those three then give only title, place and an approximate date), unless a filter needs what only the job page has: then the pages are still opened and only the text is left out.
• Workday's own count stops at 2,000; for bigger boards atsSays is the total of Workday's facet counts.
• If an ATS does not give a publish date and you set Posted in the last N days, that job is left out (it cannot be checked).
• Leave the proxy setting as it is: the ATS boards are public. If one ever blocks Apify's datacenter IPs, the run moves to residential IPs by itself.

Personal data

The rows are about jobs and companies. The Actor removes personal e-mail addresses and phone numbers from the ad text ([email removed], [phone removed]); a department's address such as careers@ or accommodations@ stays. Names that appear in an ad text stay; if you store them, data-protection law may apply to you. This Actor is not affiliated with, endorsed by or sponsored by any of the ATS vendors or companies named here; their names are trademarks of their owners.

Questions or a missing ATS?

Open an issue in the Issues tab. Answers come in writing.

Output reference

Dataset: one row per job

Field	Type	What it is
company	string	Company name as the board gives it (never the name you typed).
ats	string	greenhouse, lever, ashby, workday, smartrecruiters, recruitee, workable, personio or rippling.
board	string	The board's slug or site on that ATS.
jobId	string	The ATS's job ID.
title	string	Job title.
url	string	Link to the job.
applyUrl	string	Link to apply.
department	string or null	Department, when the ATS gives one.
team	string or null	Team, when the ATS gives one.
location	string	Main location.
locations	array of string	All locations of the job.
countryCode	string	ISO country code of the location.
workplaceType	string	REMOTE, HYBRID or ON_SITE.
employmentType	string or null	FULL_TIME, PART_TIME, CONTRACT, TEMPORARY or INTERN.
employmentTypeRaw	string or null	The ATS's own words for the employment type.
postedAt	string	First published (ISO 8601).
updatedAt	string or null	Last updated, when the ATS gives it.
salaryMin	integer or null	Salary minimum as a number.
salaryMax	integer or null	Salary maximum as a number.
salaryCurrency	string or null	ISO currency code.
salaryPeriod	string or null	e.g. YEAR, MONTH, HOUR.
salaryText	string or null	Salary as written.
salarySource	string or null	"ats" (the ATS's salary field) or "description" (the pay-range sentence of the ad).
requisitionId	string or null	Requisition ID, when the ATS gives one.
descriptionText	string	Full description as text (personal e-mails and phone numbers removed); empty with includeDescription off.
descriptionHtml	string	Full description as HTML (same removal).
contactsRemoved	integer	Personal e-mails and phone numbers removed from the text.
input	string	Your input line this job came from.

SUMMARY record (run's key-value store, never charged)

Field	Type	What it is
companies	array of object	One item per input line (fields below).
jobsDelivered	integer	Jobs delivered in the run.
onlyNewJobs	boolean	Monitoring mode on or off.
stateName	string or null	State name used in monitoring mode.
requests	object	Request counters of the run (requests, retries, refusals or blocks, failures; whether the run moved to residential IPs).
proxy	string	Proxy mode of the run.

Each item of companies:

Field	Type	What it is
input	string	Your input line.
ats, board, boardUrl	string or null	The board it matched.
detectedBy	string or null	How the board was found (link, ats:name, link on the website, name tried on each ATS).
alsoFoundOn	array of string	Other boards found for the same name.
atsSays	integer or null	The board's own job count (for Workday over 2,000: the total of its facet counts).
listed	integer	Jobs listed on the board.
filteredOut	integer	Jobs your filters dropped (not charged).
alreadyDelivered	integer	Monitoring: jobs delivered by an earlier run (not charged).
closedSinceLastRun	integer or null	Monitoring: jobs closed since the last run.
delivered	integer	Jobs delivered (and charged).
complete	boolean	true only when the whole board was read and nothing stopped it.
missing	integer or null	Gap between the board's count and what was listed.
stoppedBy	string or null	Why it stopped early (maxJobsPerCompany, your maximum charge…).
splits, incompleteParts	integer, array	Workday: smaller searches it took, and parts it could not finish.
detailsFailed, detailsFailedJobs	integer, array	Job pages that could not be read (not charged).
warnings, errors	array of string	Warnings (e.g. found by name: check the company) and errors.

More Actors by this developer

• Google Ads Transparency Scraper
• SEEK Jobs Scraper (Seek, JobStreet, Jobsdb)
• AliExpress Search Scraper
• Google News Scraper
• Sports Odds vs Prediction Markets (Kalshi, Polymarket)
• Google Trends Scraper
• YouTube Transcripts in Bulk (whole channels, playlists)

---

Versión en español

Todos los avisos abiertos de las empresas que le pases, leídos en vivo de sus propias bolsas de empleo: Greenhouse, Lever, Ashby, Workday, SmartRecruiters, Recruitee, Workable, Personio y Rippling. Le podés dar un enlace, una web o solo el nombre. Los nueve salen con la misma forma, con el sueldo como números, y cada empresa termina con un control: cuántos avisos dice su bolsa que tiene y cuántos recibiste.

De un vistazo

• Qué hace: Todos los avisos abiertos de las empresas que le pases, leídos en vivo de sus propias bolsas en 9 sistemas (Greenhouse, Lever, Ashby, Workday, SmartRecruiters, Recruitee, Workable, Personio, Rippling), con la misma forma de fila, sueldo como números y Workday más allá de su tope de 2.000.

• Sirve para: seguir los avisos de una lista de empresas; señales de contratación; feeds sin el atraso de una base de datos; monitorear solo avisos nuevos.

• No sirve para: portales que juntan muchas empresas (Indeed, LinkedIn); BambooHR, Teamtailor, Breezy, iCIMS, Oracle o SuccessFactors (no están en esta versión); encontrar un Workday solo por nombre.

• Entrada: companies (obligatorio): enlaces de bolsa, ats:nombre, la web o el nombre; Workday necesita su enlace. Filtros opcionales (antes de cobrar), maxJobsPerCompany, onlyNewJobs + stateName, includeDescription. Ejemplo mínimo en JSON: sección Minimal input de la parte en inglés.

• Salida: una fila por aviso y un registro SUMMARY con "la bolsa dice N, entregados N" por empresa.

• Eventos que se cobran: job USD 0,0015 por aviso entregado (USD 1,50 cada 1.000), con el texto completo; más apify-actor-start.

• Nunca se cobra: las empresas en sí (no hay cargo por empresa), una bolsa que no se encuentra, avisos que tus filtros descartan, avisos ilegibles, avisos ya entregados en modo monitoreo, el SUMMARY.

• Límites: se detiene en maxJobsPerCompany o en tu costo máximo por corrida (stoppedBy); si la bolsa se encontró por nombre, SUMMARY lo dice (detectedBy).

• En vivo, no una base de datos: lo publicado hace diez minutos está; lo cerrado, no.

• Workday pasando su tope de 2.000: Workday nunca lista más de 2.000 avisos por búsqueda; el programa divide la búsqueda por los filtros del propio Workday y junta las partes. NVIDIA: 2.680 de 2.680 (otro programa de la tienda, el mismo día: 2.000); Target: 14.130 de 14.149.

• Sueldo como números (mínimo, máximo, moneda, período), del campo de sueldo del sistema o, si no tiene, de la frase del rango salarial del aviso.

• Solo avisos nuevos: corriéndolo cada día con la misma lista, recibís (y pagás) solo lo que no recibiste antes, más cuántos avisos se cerraron.

Qué le das: enlaces de las bolsas (o de un aviso), ats:nombre (greenhouse:stripe), la web de la empresa (figma.com) o el nombre (Datadog). Si se encontró por nombre, el resumen lo dice y pide revisar que sea la empresa correcta. Workday necesita su enlace.

Filtros (pagás solo por lo que pasa): palabras del puesto (incluir/excluir), lugar o país, modalidad, área, publicados en los últimos N días, máximo por empresa.

El registro SUMMARY dice, por cada línea: qué bolsa se usó y cómo se encontró, cuántos avisos dice la bolsa, cuántos se listaron, filtraron y entregaron, y por qué no está completa si no lo está.

Precio: USD 1,50 cada 1.000 avisos entregados, con el texto completo. Nada por empresa, nada si no se encuentra la bolsa, nada por los avisos que tus filtros descartan. Más el evento de inicio estándar de Apify (USD 0,00005 por corrida, por GB de memoria). Se detiene en tu costo máximo por corrida y lo informa en SUMMARY.

Datos personales: se borran los correos y teléfonos personales del texto del aviso (los de un área, como careers@, quedan). Este programa no está afiliado a ninguno de los sistemas ni empresas nombrados.

¿Dudas o falta un sistema? Abrí un issue en la pestaña Issues; se responde por escrito.