Job Search Scraper — Live Jobs Direct from Company Career Pages

Search live job listings read straight from company career pages, across 2,335 verified employers on 17 ATS platforms: Greenhouse, Workday, Lever, Ashby, SmartRecruiters, SAP SuccessFactors, Oracle Cloud HCM, Phenom, Eightfold and more. Powered by the spool job engine. No job-board reposts, no expired or ghost jobs, no proxies, no login.

Senior Software Engineer, Payments     Stripe        Remote, US        $180,000–$260,000   Software Engineering · senior   greenhouse
Staff Machine Learning Engineer        Databricks    San Francisco     $190,000–$285,000   AI & Machine Learning · lead    greenhouse
Lead, Systems Engineering              L3Harris      Ottawa, CA        $126,500–$176,500   Manufacturing & Hardware        successfactors
Security Engineer, Cloud               Ramp          New York (HQ)     $175,000–$220,000   Security · mid                  ashby

Why this scraper instead of scraping LinkedIn or Indeed?

• Direct from the source. Every posting comes from the employer's own career page, read at query time. What you get is what the employer shows today, with the real posting date and the direct application link.
• No expired or ghost jobs. Nothing is served from a stale database. A posting that is gone from the career page cannot appear in your results.
• No duplicate spam. One canonical listing per role. Set the deduplication mode to collapse an employer's multi-location copies into one row.
• Enriched fields on every row. Job category (30), employment type, experience level and years asked, visa sponsorship, education, 190 technologies as skills, salary annualised in USD, parsed city, region and country, remote or hybrid or on-site.
• No proxies. The boards publish public JSON meant for exactly this, so runs are cheap and start fast: the first rows land within seconds, the default pool of 400 employers finishes in about 30–60 seconds; a sweep of the whole list of 2,335 takes a few minutes (set employerPool to 2335).

What you can filter by

Filter	Examples
Job titles	Software Engineer, Account Executive, Data Analyst (phrase or every word in the title)
Keywords	Python, Kubernetes, HIPAA, Salesforce — in the title, skills or posting text, any of them
Job categories	30 categories, from Software Engineering and AI & Machine Learning to Healthcare and Skilled Trades
Workplace type	Remote, hybrid, on-site
Location	New York · Texas · London · Germany, matched against the employer's label and the parsed city, region, country
Countries	Two-letter codes: US, CA, GB, DE, IN
Employment type	Full-time, part-time, contract, internship, temporary
Experience	junior → executive, plus minimum and maximum years of experience asked
Posted	Within the last N days
ATS platform	Greenhouse, Workday, Lever, Ashby and 13 more
Salary	Only postings with a stated salary; a USD floor, annualised
Visa sponsorship	Only postings whose text says sponsorship is available
Technologies	Only postings naming any of these (190 recognised)
Companies	Add your own employers by name, domain or careers URL; they are found through their careers page if needed. When they leave room under your maximum, the same kinds of roles at comparable employers fill the rest, each row marked similar
Extras	Deduplicate by company and title, only new since last run, up to 5,000 results per run

Output

Each result carries: job title, company and website, direct application link, ATS platform and board URL, a stable jobSlug for syncing, posting date and days since posting, location with parsed city, region, country and workplace type, employment type, job category, department, experience level and years of experience, education level, visa sponsorship (true, false or not stated), salary as stated and annualised in USD, skills and skill categories, the employer's own description as clean text, and the title or keyword that matched.

{
  "jobTitle": "Senior Software Engineer, Data Infrastructure",
  "company": "Stripe",
  "companyWebsite": "stripe.com",
  "applicationLink": "https://stripe.com/jobs/listing/senior-software-engineer-data-infrastructure/6198542",
  "jobBoard": "greenhouse",
  "datePosted": "2026-10-03",
  "location": "Toronto, Canada",
  "locationType": "hybrid",
  "country": "CA",
  "employmentType": "fulltime",
  "jobCategory": "Software Engineering",
  "experienceLevel": "senior",
  "yearsExperienceMin": 5,
  "visaSponsorship": null,
  "salaryMinUsd": 164000,
  "salaryMaxUsd": 246000,
  "skills": ["Python", "Kafka", "Kubernetes", "AWS"]
}

Export to JSON, CSV or Excel, or feed it into Make, n8n, Zapier, Google Sheets or the Apify API. The Jobs, Salaries in USD and Experience, visa, education views on the Output tab each show their own columns.

Example searches

Job seeker, a daily remote digest. Job titles Software Engineer, workplace type remote, experience level senior, minimum salary 150000, posted within 7 days. Run it on an Apify Schedule every morning with Only new since last run on and add the Gmail or Slack integration: new matches arrive in your inbox, nothing twice.

{ "jobTitles": ["software engineer"], "workplaceType": ["remote"], "experienceLevel": ["senior"], "minSalaryUsd": 150000, "postedWithinDays": 7, "onlyNew": true }

Recruiter, map a niche talent market. Job category Sales, keywords fintech, employer pool 800. Export to Excel and you have every live sales opening at the verified employers that mention fintech, with salaries where stated.

{ "jobCategories": ["Sales"], "keywords": ["fintech"], "employerPool": 800 }

Sponsorship search. Keywords Kubernetes, visa sponsorship on, countries US. Only postings whose text says sponsorship is available.

{ "keywords": ["Kubernetes"], "visaSponsorship": true, "countries": ["US"] }

Data team, keep a job board fresh. One category per scheduled run, maxResults 5000, deduplicate by company and title, only new since last run. Insert by jobSlug; a posting that disappears from a career page simply stops coming back.

{ "jobCategories": ["Data & Analytics"], "maxResults": 5000, "dedupeMode": "companyTitle", "onlyNew": true, "employerPool": 800 }

Your own watch-list, searched the same way. Add companies by name, domain or careers URL; they are found through their careers page when the name alone is not enough. A short result is filled with the same kinds of roles at employers like them: same industry, nearest team size, from the verified list. Every added row says matchType: similar and names the company it is closest to in similarTo, and a notice at the end says how many were added. Set similarCompanies to 0 for the named companies only.

{ "jobTitles": ["account executive"], "companies": ["Datadog", "stripe.com", "https://jobs.ashbyhq.com/ramp"], "employerPool": 0 }

One company, and who else hires the same roles. Name a company with no title and the run returns its whole board, then the same categories of roles at its peers until the maximum is reached.

{ "companies": ["Ramp"], "employerPool": 0, "maxResults": 1000, "similarCompanies": 100 }

Measured on the platform, 11 October 2026 (builds 1.0.6 to 1.0.8). Every row below is one real run; costs are the per-posting charge, before the Actor start.

Search	Boards	Postings scanned	Delivered	First rows	Done	Cost
Default: remote software engineers, last 30 days	300	60,818	285 (217 with salary, 261 with description)	1.8 s	3 min 45 s	$0.86
Three named companies, account executive and SDR titles	3	1,328	100	1.8 s	2.2 s	$0.30
Keyword Kubernetes, stated salary only	32	18,777	150	98 s	98 s	$0.45

Company searches with the fill, measured 29 October 2026 (build 1.0.11).

Search	Named company rows	Added similar rows	Done	Cost
Ramp, no title, maxResults 1000	168	832 at 88 employers of comparable size, in the categories Ramp hires for	2 min 32 s	$3.00
Stripe, no title, maxResults 1000	707	293 at 86 employers, 75 of them Fintech	1 min 23 s	$3.00
Datadog, account executive, maxResults 300	47	253 account executive roles at 59 employers	64 s	$0.90
Ramp, similarCompanies 0	168	0	4 s	$0.50
US tech roles offering visa sponsorship (cap 60)	250	272,781 listed	57, all with the sponsorship sentence	78 s
Remote roles posted in the last 24 hours (cap 60)	250	285,419 listed	7, all dated today or yesterday	12 s
Software, data and AI internships (cap 60)	250	18,106	60, all internships	6 s
Entry-level software engineers, two years or less (cap 60)	250	18,506	60	3.5 s
Workday employers only (cap 60)	82	6,400	60, all Workday	4.7 s
Tech roles paying $200k or more (cap 60)	250	28,481	65, every one at or above the floor	58 s

Searches that depend on the posting text (keywords, visa sponsorship, salary floors, years of experience) read every candidate's description and take longer; title, category, location, platform and date searches are answered from the listings and start within seconds.

Ready-made searches

Each of these is a public task with its own page and schedule button. Open one, press Start, or save it as your own and change the filters.

Search	What it returns
Remote senior software engineer jobs	Remote, senior, posted in the last 7 days
Remote jobs posted today	Every remote role published in the last 24 hours, the whole list
Visa sponsorship tech jobs USA	US tech roles whose text says sponsorship is available
Tech jobs paying over $200k	Stated pay of $200,000 a year or more, annualised in USD
Entry-level software engineer jobs	Junior roles asking for two years of experience or less
Software engineering internships	Internships and co-ops in software, data and AI
Remote data engineer jobs	Data and analytics engineers, posted in 14 days
Machine learning engineer jobs	The AI and machine learning category, 30 days
Product manager jobs	Product management roles, 30 days
Workday jobs	Every Workday employer in the verified list

Input

Field	Type	What it does
jobTitles	list	Keep postings whose title contains the phrase or every word of it
keywords	list	Keep postings naming any of these in the title, skills or text
jobCategories	list	One or more of the 30 categories
workplaceType	list	remote, hybrid, onsite
location	text	Location label, city, region or country contains this text
countries	list	Two-letter codes
employmentType	list	fulltime, parttime, contract, internship, temporary
experienceLevel	list	junior, mid, senior, manager, director, executive
minYearsExperience, maxYearsExperience	number	Against the years the posting asks for; postings stating no number are kept
postedWithinDays	number	Employer's posting or update date within N days
atsPlatforms	list	Only employers on these systems
requireSalary, minSalaryUsd	flag, number	Stated salary only; USD floor, annualised
visaSponsorship	flag	Only postings whose text says sponsorship is available
requireTech	list	Only postings naming any of these technologies
companies	list	Extra employers by name, domain or careers URL, up to 100
similarCompanies	number	When the named companies leave room under maxResults, add the same kinds of roles from up to this many comparable employers (default 100, 0 turns it off)
employerPool	number	How many verified employers to search, largest first (default 300, up to the whole list)
maxResults	number	Stop after this many postings (default 500, up to 5,000)
dedupeMode	choice	location (default), companyTitle, none
maxDescriptionsPerBoard	number	Descriptions per Workday, Oracle, SuccessFactors, Phenom or Eightfold board
onlyNew, monitorName	flag, text	Only postings not delivered before; the memory record's name

Output fields

Field	Meaning
jobTitle, company, companyWebsite, companyIndustry, companySize	The posting's title and the employer, with industry and team size where the verified list knows them
applicationLink, jobBoard, boardUrl, jobId, jobSlug	Direct apply link, the ATS platform, the board, the platform's id, and a stable key for syncing
datePosted, postedDaysAgo	The employer's own date
location, city, region, country, locationType, isRemote	As stated, then parsed; remote, hybrid or onsite
employmentType, jobCategory, department, experienceLevel	From the platform's field where it has one, otherwise from the title and text
yearsExperienceMin, yearsExperienceMax, educationLevel, visaSponsorship	Read from the posting text; visaSponsorship is true, false or null
salaryMin, salaryMax, salaryCurrency, salaryPeriod, salaryMinUsd, salaryMaxUsd, salarySource, hasSalary	As stated, then annualised in USD; ats or description
skills, skillCategories	Technologies named in the posting, 190 recognised
descriptionText, descriptionChars, matchedOn, dedupeKey, firstSeenAt	The employer's text, what matched, and the only-new bookkeeping
matchType, matchedOn, similarTo	company for an employer you named, verified for the verified list, keyword for an engineering role whose posting text names the technology in your title (the word in matchedOn), similar for a comparable employer added to fill the result, with the named company it is closest to

Live search or jobs database?

	This Actor	A jobs database API
Where results come from	The employer's career page, read when you run it	A crawl stored hours or days ago
Expired and ghost jobs	Cannot appear	Removed when the next re-check finds them gone
Coverage	2,335 verified employers plus any company you name	Whatever the database crawled
Time to results	First rows in seconds; 300 employers in two to four minutes	Seconds
Proxies	None	Not your concern
Best for	Feeds that must be exact, watch-lists, company-level signals	Broad keyword search across the whole market

Where the jobs come from

The search runs live over a bundled, verified list of 2,335 employers whose public boards answered when the list was built: Greenhouse (516), Ashby (498), Workday (488), SAP SuccessFactors (141), Lever (122), Phenom (112), Oracle Cloud HCM (108), Pinpoint (66), Breezy (52), Rippling (50), Personio (48), Teamtailor (44), SmartRecruiters (33), BambooHR (28), Eightfold (25), Recruitee (3), Amazon (1). Together they hold about 353,000 open roles on any given day, and 610 of the companies carry their industry and team size from the YC company data, from Stripe, Databricks, OpenAI and Anthropic to Lockheed Martin, Capital One, JPMorgan Chase, Nvidia, ExxonMobil and L3Harris. employerPool sets how many of them to search, largest first; companies adds any employer you name.

The list is rebuilt from public seeds: every US-listed company (the SEC's register), the S&P 500, 2,000 YC companies, US hospitals (the CMS register) and US universities, and grows with every pass. Companies you add are resolved by the same engine: name → board slug on every platform at once → the company's careers page and the system behind it, within six requests and ten seconds.

Pricing

$3.00 per 1,000 postings, on every plan. Pay per result: you are charged for postings actually delivered, up to 5,000 per run, plus Apify's small Actor-start fee. Duplicates, postings removed by your filters and postings already delivered in only-new mode cost nothing.

Search	Postings	Cost
Default run (remote software engineers, last 30 days, 300 employers)	285	$0.86
1,000 postings	1,000	$3.00
A category across the whole list, 5,000 cap	up to 5,000	up to $15.00
Daily only-new feed on a quiet day	0	the Actor start

The free plan's $5 monthly credit covers about 1,600 postings.

Use it from an AI agent, the API or an integration

MCP. Add the Apify MCP server to Claude, Cursor or any MCP client and the agent can call this Actor as a tool: "Find remote senior data engineering roles posted this week that pay over $180k and mention Snowflake."

https://mcp.apify.com/?tools=spool/job-search

API. POST https://api.apify.com/v2/acts/spool~job-search/run-sync-get-dataset-items?token=… with the filters as JSON returns the rows in one call; an empty body runs the default search. Python and JavaScript: client.actor("spool/job-search").call(run_input={...}) with apify-client. Schedules and integrations: run daily with onlyNew and push rows to Google Sheets, Airtable, Slack, HubSpot or a webhook from the Integrations tab.

FAQ

Do I need an account or API key for the job boards? No. The boards publish public listings; the Actor reads them directly. You pay per result through Apify only.

How is this different from a jobs database API? A database serves what it crawled hours or days ago and must re-check listings to remove the expired ones. This Actor reads the career pages when you run it, so the results are the employer's live board. The trade-off is time: rows start within seconds, a search over the default 400 employers finishes in about a minute, and the whole list takes a few minutes.

Why do I sometimes see the same job more than once? Some employers publish one posting per location. Set Deduplicate results to companyTitle to collapse them; skipped duplicates are not billed.

Can I run it on a schedule and get only new jobs? Yes. Turn on Only new since last run: the search remembers every posting it delivered, in a key-value store in your own account, and returns only new ones next time. The first run returns normal results and records the starting point. Changing the filters starts a fresh memory unless you set a monitorName.

What does "visa sponsorship" mean here? true when the posting text says sponsorship is available, false when it says it is not, null when it says nothing. The filter keeps only true.

How fresh are the dates? datePosted is the employer's own posting or last-update date. Where a platform gives no date, the field is empty and the posting is kept by the date filter rather than silently dropped.

I searched "python developer" and got "Software Engineer" rows. Employers rarely put the language in the title; the posting text names it. When a title names a technology, the run adds engineering roles whose text names it, marked matchType: keyword with the word in matchedOn, after the title matches. Filter them out by matchType if you want titles only, or set keywords to search the text on your own terms.

I named one company and got rows from others. That is the fill: when the companies you named leave room under maxResults, the run adds the same kinds of roles at comparable employers, chosen by industry and team size from the verified list. Those rows are marked matchType: similar, similarTo names the company they are closest to, the notice row says how many were added, and RUN_SUMMARY.similar lists the basis. Set similarCompanies to 0 to receive only the companies you named; you pay only for rows delivered either way.

A company I added was not found. RUN_SUMMARY.companies.notFound says why: the careers page uses a system the engine cannot read yet (iCIMS, Paradox, Taleo, Avature, ADP, UKG are recognised and named), the website did not answer, or no careers page was found. Paste the careers URL and the engine reads what is behind it.

Is this legal? The data comes from job listings employers publish openly on their own career pages through public endpoints intended for exactly this. No login, no candidate data, no paywall bypass. Handle any personal data a description happens to contain according to the law that applies to you.

Changelog

1.4.1 — 2026-11-03. Keyword rows are exact: the technology must appear as its own word, so "React" no longer matches "reactive" or "reaction" (36 of 191 React rows in a measured run), and iOS keeps its case, so a "Cisco IOS" network posting is not an iOS row. Keyword rows come in two tiers: roles of the kind your titles name first ("python developer" wants engineers, developers, programmers and architects; "salesforce administrator" wants administrators, consultants and specialists), then adjacent roles whose posting names the technology, only to fill what is left; RUN_SUMMARY.keywordPass.roles gives the split. "Engineering Manager" and "Head of Engineering" no longer match an engineer search through the word "engineering".

1.4.0 — 2026-11-03. Speed and a keyword pass. A default run of 400 employers now finishes in about 30–60 seconds (it was four minutes): every board is read through its own small connection pool, enterprise boards (Workday, Oracle, Phenom, SuccessFactors) run as a parallel stream beside the fast boards instead of after them, posting text is stripped only for rows that are delivered, and the default memory is 4 GB. When a job title names a technology (Python, React, Salesforce, iOS, Java, EDI, SAP, Kubernetes …), a second pass searches the same boards for engineering roles whose posting text names that technology, because most employers title the role "Software Engineer" and name the language in the text. Those rows are marked matchType: keyword with the word in matchedOn, come after the title matches, are spread across employers, and a notice reports how many were added (RUN_SUMMARY.keywordPass). Set keywords yourself to search the text directly instead.

1.3.0 — 2026-11-03. Coverage. The verified list grows to 2,335 employers (353,000 open roles) from new public seeds: every US-listed company in the SEC register, US hospital systems from the CMS register and US universities, resolved through their careers pages to 17 platforms (Greenhouse (516), Ashby (498), Workday (488), SAP SuccessFactors (141), Lever (122), Phenom (112), Oracle Cloud HCM (108), Pinpoint (66), Breezy (52), Rippling (50), Personio (48), Teamtailor (44), SmartRecruiters (33), BambooHR (28), Eightfold (25), Recruitee (3), Amazon (1)). Greenhouse boards now deliver up to 150 descriptions per board in a pool search (was 30, which left a fifth of default-run rows without text). employerPool can be set to the whole list. Salaries stated on an employer's board are annualised in USD on every platform (1.2.1).

1.2.1 — 2026-10-31. Salaries stated on the employer's own board (Ashby and others send the pay period as "1 YEAR" or "1 HOUR") are now annualised and converted to USD like every other salary. Before this fix those rows carried the pay in salaryMin/salaryMax but no salaryMinUsd/salaryMaxUsd, the minSalaryUsd filter dropped them, and the Salaries in USD view left them out. Found in a quality pass: 171 of Ramp's 191 salaried roles, 86 of 150 Kubernetes roles with stated pay.

1.2.0 — 2026-10-29. Company searches fill up: when the companies you name leave room under maxResults, the same kinds of roles at comparable employers (same industry, nearest team size, from the verified list) are added, up to similarCompanies boards (default 100). Every row now carries matchType (company, verified or similar) and similarTo; a notice reports what was added and RUN_SUMMARY.similar the basis. Industry and team size are now attached to named companies that appear in the verified list.

1.1.0 — 2026-10-11. The verified list grows to 1,129 employers (68,000 open roles) after a pass over 1,475 more YC companies; 610 of them carry industry and team size, now on every row as companyIndustry and companySize. Employment type comes from the platform's own field on Ashby, Lever, Workday and SmartRecruiters. Category, seniority, employment and date filters now run before descriptions are fetched, so filter-only searches finish in a fraction of the time; filters that need the posting text (keywords, visa sponsorship, years, salary floor, technologies) read every candidate's description. Date filters only return dated postings; Workday roles get their date from the listing. Live progress messages during the run.

1.0.0 — 2026-10-11. First release: live search over 768 verified employers on 17 ATS platforms plus any company you name; filters for titles, keywords, 30 categories, workplace type, location and country, employment type, experience level and years, posting date, platform, salary, visa sponsorship and technologies; deduplication modes; only-new mode; rows delivered in waves as boards answer.

More from spool

• Career Page Scraper & ATS Jobs API — 17 ATS Platforms, No Proxy — whole boards for the companies you name, with hiring profiles per company
• Job Board Scraper & API — LinkedIn, Indeed, Glassdoor + 29 More — one search across 33 sources with the employer's posting attached
• Company Tech Stack API — Detected from Job Postings — a company's technologies read from its own job board