Multi-ATS Job Board Scraper (Greenhouse, Lever, Ashby, SmartRecruiters, Recruitee)

Fetch normalized job postings from five popular ATS platforms — Greenhouse, Lever, Ashby, SmartRecruiters, and Recruitee — in a single run. Get title, location, department, employment type, description, and apply URL, ready for pipelines, dashboards, or AI search.

Why use this actor

Talent data is siloed across ATS platforms. A single company may post jobs on Greenhouse while a competitor uses Lever. Checking each platform manually, or building and maintaining five separate API integrations, is expensive.

This actor gives you one normalized job dataset from all five platforms with a single input. It's designed for:

• Recruiters and sourcing teams monitoring competitor or partner job boards
• Product managers analyzing hiring trends by department, location, or role
• Data engineers building talent-intel pipelines or company-growth signals
• Developers automating job alerts, career aggregators, or internal hiring dashboards
• VC/PE analysts tracking portfolio company headcount via live job postings

Supported platforms

Platform	Input example	Notes
Greenhouse	{ "platform": "greenhouse", "companyId": "vercel" }	Official JSON board API; per-job description fetch available
Lever	{ "platform": "lever", "companyId": "netflix" }	Official v0 postings API; crawl-delay honored
Ashby	{ "platform": "ashby", "companyId": "notion" }	Documented public posting-api endpoint
SmartRecruiters	{ "platform": "smartrecruiters", "companyId": "McDonalds" }	Official /v1/companies API
Recruitee	{ "platform": "recruitee", "companyId": "acme" }	Public /api/offers/ endpoint

Input

Field	Type	Default	Description
sources	array	required	List of {platform, companyId} objects
maxJobsPerCompany	integer	100	Max jobs per source, capped at 500
includeDescription	boolean	true	Fetch and include full job description (plain text, HTML stripped)
scrubPii	boolean	true	Redact incidental email/phone in descriptions (GDPR)
maxTotalChargeUsd	number	10.0	Hard charge cap; actor stops gracefully when reached

Output

One dataset record per job posting. Fields:

{
  "platform": "greenhouse",
  "companyId": "vercel",
  "jobId": "7612345",
  "title": "Senior Software Engineer, Edge Network",
  "location": "Remote",
  "department": "Engineering",
  "employmentType": null,
  "description": "Join Vercel's Edge Network team...",
  "applyUrl": "https://boards.greenhouse.io/vercel/jobs/7612345",
  "postedAt": null,
  "updatedAt": "2026-06-15T10:30:00Z",
  "fetchedAt": "2026-07-28T12:00:00Z"
}

Error records (invalid companyId, platform unreachable, zero results) are pushed to the dataset with "status": "error" and a plain-English "reason", so the run never exits silently on bad input.

Worked example + cost math

Input:

{
  "sources": [
    { "platform": "greenhouse", "companyId": "vercel" },
    { "platform": "lever", "companyId": "netflix" },
    { "platform": "ashby", "companyId": "notion" }
  ],
  "maxJobsPerCompany": 50,
  "includeDescription": true
}

Expected events and cost:

Event	Count	Unit price	Subtotal
actor-start	1	$0.01	$0.01
job-fetched	~150 (50 per co.)	$0.003	$0.45
Total			~$0.46

For 3 companies × 50 jobs each (150 jobs):

• Platform compute: 0.005 CU × $0.40/CU = **$0.002** (< 1% of revenue — well under the 30% cost cap)
• Revenue per run: ~$0.46
• Gross margin: ~99%

At $0.003 per job, a user fetching 1,000 jobs/month spends $3.00 — comparable to a manual analyst spending 2–3 hours.

Compliance and rate limiting

• All five platforms are accessed via their official, documented public APIs (not HTML scraping)
• robots.txt honored on all target domains (Greenhouse: allows all; Lever: crawl-delay 1 honored; Ashby posting-api: not disallowed; SmartRecruiters /v1/companies/: explicitly allowed; Recruitee: Cloudflare-gated but public)
• Rate limited to ~1 request/second per-tenant, well within each platform's tolerance
• Descriptive User-Agent header set on every request
• Incidental PII (recruiter emails/phones in descriptions) is redacted by default (scrubPii: true)
• No login-walled content, no candidate data, no session cookies

Limitations

• Workday is not supported: Cloudflare + Akamai bot-management makes reliable access infeasible without a browser; a separate Workday actor is warranted
• SmartRecruiters descriptions require a per-job secondary fetch that would multiply requests; descriptions are omitted for that platform to stay polite
• Recruitee descriptions similarly require a per-posting HTML parse; omitted for the same reason
• Company IDs are platform-specific slugs — use the slug visible in the platform's job board URL (e.g. boards.greenhouse.io/vercel → vercel)

Integrations

• Apify API: POST /v2/acts/gp005~multi-ats-jobs-scraper/runs — get a JSON feed of jobs in any pipeline
• Make / n8n: Trigger on schedule, filter by department or location, post to Slack/Sheets
• MCP: Available via the Apify MCP server (apify/actors-mcp-server) — let your AI assistant query live job postings
• Webhooks: Use ACTOR_RUN_FINISHED webhooks to push new jobs to your CRM, ATS, or data warehouse

FAQ

Which companyId should I use? Use the slug in the platform's job-board URL. For Greenhouse it's boards.greenhouse.io/<slug>, for Lever it's jobs.lever.co/<slug>, for Ashby it's jobs.ashbyhq.com/<slug>.

Can I monitor new postings automatically? Yes — schedule the actor on a daily/weekly cadence and compare datasets across runs, or use Apify webhooks to get notified on completion.

Does this work for private/internal job boards? No — only publicly accessible boards (no login required) are supported.

Is this GDPR-compliant? The core data (job titles, departments, locations) is not personal data. Descriptions may incidentally include recruiter contact details; the scrubPii: true default redacts those. Users are responsible for their own downstream data handling.

What if a company has no open jobs? An error record is pushed with "reason": "No job postings found..." so you always get a result, not a silent empty run.