Hiring intelligence · ATS · July 21, 2026
20,487 discovered company career-site boards across 14 ATS providers — and 624,299 live postings crawled from them, normalized into one JSON shape with company, department, location, employment type and remote flag. Search across every provider, or find which companies are hiring right now. Public postings only, pay on success.
20,487
company boards tracked — 14 ATS providers, discovered and crawled.
70.3%
active boards
17.1%
blocked or gone
624K
live postings
Snapshot July 21, 2026 — the crawl updates continuously; most recently seen postings are same-day.
20,487
discovered company career-site boards across 14 ATS providers, carrying 624,299 live postings. Workday, Greenhouse and Workable lead by board count; iCIMS and Eightfold were added most recently and are still ramping.
17.1%
of tracked boards are currently gone or actively blocking the crawl — public career pages churn constantly: companies switch ATS providers, retire boards, or add bot protection. A board registry needs continuous re-discovery, not a one-time seed list.
11%
of postings carry an explicit remote / workplace-type flag — that is a floor on remote postings, not the true remote share. Most ATS boards never expose the field at all, so a role can be fully remote without ever being flagged as such.
7+
distinct raw spellings for one concept — full-time employment — are seen across the 14 providers' employment_type field. Normalizing that mess is most of the value of querying one dataset instead of scraping fourteen ATS layouts.
Workday, Greenhouse and Workable together account for well over half of every discovered board; Lever, iCIMS and Eightfold are the long tail.
iCIMS and Eightfold were only just added to the crawl, so their board counts are not yet comparable to the other 12 providers.
Every discovered board carries a lifecycle status. Only just over half are currently active and returning postings — the rest are blocked (rate-limited or challenged), gone (the company left that ATS or the board 404s), empty (a valid board with zero open roles), or freshly discovered and not yet crawled.
Only postings where the provider tags a department participate in this breakdown — several boards never populate the field at all, so treat this as a partial-coverage signal, not a department census of every open role.
A sample of companies with a large number of open roles in the latest snapshot, deliberately shown without a headline count: several land on suspiciously round numbers — exactly 2,000 or 4,000 open roles — which almost certainly reflects a per-board crawl page cap rather than a literal count. Query the API directly for the current figure on any one company.
Company names only — see the caveat above before treating any open-role total as exact.
Two record types: one board per discovered company career-site, one posting per crawled job. Grouped for readability; the API returns flat objects.
Board record
Board identity
board_idprovidercompany_namedomaincareers_urlBoard status
statusopen_roles_countdiscovery_sourcefirst_seenlast_crawledlast_errorPosting record
Posting identity
posting_idboard_idprovideridtitlecompanyPosting classification
departmentteamlocationlocations[]remoteworkplace_typeemployment_typeCompensation
compensationcompensation_mincompensation_maxcompensation_currencyFreshness & linkage
urlapply_urlreq_idposted_atposted_age_textfirst_seenlast_seenis_openTech-stack hint (company detail, when a domain is known)
domaintechnology_names[]categories[]cmsecommercecdnTwo search surfaces: search postings directly, or search the board registry to see which companies are hiring, on which ATS, with how many open roles. Page with page and page_size (search: page × page_size ≤ 10,000).
Posting search
Company / board search
Geo radius search
Facet aggregation
Sort: relevance, posted_desc, company_asc (search) · open_desc, company_asc, crawled_desc (companies)
Every query authenticates with an x-api-key header and reads the stored search index — there is no live crawl to trigger per request, and you are billed pay on success: charged for results, not failed requests. Six endpoints cover it:
GET /datasets/jobs/search — filter, sort, page every crawled posting.GET /datasets/jobs/items/{id} — one posting by id.GET /datasets/jobs/facets — provider, department, employment-type and remote aggregates in one call.GET /datasets/jobs/nearby — postings within a radius of a coordinate.GET /datasets/jobs/companies — which companies are hiring, on which ATS, with how many open roles.GET /datasets/jobs/companies/{id} — one company board by id, with a tech-stack hint when a domain is known.Five of the six are exposed as MCP tools — datasets_jobs_search, datasets_jobs_companies, datasets_jobs_facets, datasets_jobs_item and datasets_jobs_nearby — and covered by all six client SDKs, so an agent can call them directly.
Cite this
Crawlora (2026). Jobs Dataset. 20,487 company boards, 624,299 postings; public data only. https://crawlora.net/datasets/jobs.
Search postings
# Remote engineering roles, most recently posted first
curl "https://api.crawlora.net/api/v1/datasets/jobs/search?department=Engineering&remote=true&sort=posted_desc" \
-H "x-api-key: $CRAWLORA_API_KEY"Find who's hiring
# Which Greenhouse companies are actively hiring, most open roles first
curl "https://api.crawlora.net/api/v1/datasets/jobs/companies?provider=greenhouse&status=active&sort=open_desc" \
-H "x-api-key: $CRAWLORA_API_KEY"Facet the market
# Provider, department, employment-type and remote aggregates in one call
curl "https://api.crawlora.net/api/v1/datasets/jobs/facets" -H "x-api-key: $CRAWLORA_API_KEY"The same figures behind the bars above — plain and machine-readable for search engines and AI answer engines that cannot parse a chart.
| Provider | Boards | Share |
|---|---|---|
| Workday | 4,028 | 19.7% |
| Greenhouse | 3,764 | 18.4% |
| Workable | 2,980 | 14.5% |
| Ashby | 2,618 | 12.8% |
| Personio | 1,339 | 6.5% |
| UKG | 1,233 | 6% |
| Teamtailor | 1,100 | 5.4% |
| Rippling | 927 | 4.5% |
| SmartRecruiters | 893 | 4.4% |
| Oracle | 741 | 3.6% |
| Recruitee | 734 | 3.6% |
| Lever | 119 | 0.6% |
| iCIMS | 9 | 0% |
| Eightfold | 2 | 0% |
| Status | Boards | Share |
|---|---|---|
| Active | 14,408 | 70.3% |
| Blocked | 277 | 1.4% |
| Gone | 3,219 | 15.7% |
| Empty | 2,582 | 12.6% |
| Pending | 1 | 0% |
| Department | Postings | Share of tagged |
|---|---|---|
| Engineering | 16,237 | 28.5% |
| Sales | 12,427 | 21.8% |
| Operations | 6,772 | 11.9% |
| Store Hourly Positions | 5,049 | 8.9% |
| Marketing | 4,744 | 8.3% |
| Product | 3,194 | 5.6% |
| Finance | 2,938 | 5.2% |
| Government Services | 1,911 | 3.4% |
| Technology | 1,882 | 3.3% |
| Clinical | 1,845 | 3.2% |
Query 20,487 company boards and 624,299 postings — provider, department, location, employment type and remote flag — through one REST API. Company research, hiring-signal tracking, recruiting tooling or talent-market analysis: clean JSON, public data only, pay on success.
The dataset tracks 20,487 discovered company career-site boards across 14 ATS providers (Greenhouse, Lever, Ashby, Workday, SmartRecruiters, Workable, Recruitee, Rippling, Personio, Teamtailor, Oracle, UKG, iCIMS, Eightfold), carrying 624,299 live job postings. Board discovery runs continuously — a top-1M-domain ATS fingerprint scan, Common Crawl URL-index harvesting, and company-seed probing (SEC, Crunchbase, Google Business, GitHub) — so both totals grow over time.
Every discovered board carries a lifecycle status: active (last crawl returned postings), empty (a valid board with zero open roles), gone (404, redirected, or the company left that ATS), blocked (rate-limited or challenged on the last attempt), or pending (discovered but not yet crawled). About 17.1% of tracked boards are currently blocked or gone — public career pages change ATS providers, retire old boards, and add bot protection constantly, so this is an accurate picture of real-world churn, not a data-quality problem.
Treat it as a floor, not a census. 11% of open postings (68,550) currently carry remote=true after structured ATS flags plus free-text inference from location, title, workplace type, and office names. Many enterprise boards never expose a remote boolean at all, so the flag was historically under-counting Workday and Greenhouse volume; residual gaps remain (topic titles like “Remote Monitoring”, office-name-only remotes on some boards). Use location text as a secondary signal and never treat remote_open as market share.
Because it is raw from each ATS, not yet normalized on our side: at least 7 distinct spellings map to what is materially one concept, full-time employment — for example Full-time, FullTime, full-time, fulltime_permanent and full-or-part-time all show up as separate raw strings across the 14 providers. That is exactly the kind of cross-provider inconsistency a normalized dataset is meant to save you from parsing yourself.
Yes. The dataset is queryable over Crawlora's REST API and via MCP: full-text posting search, one posting by id, provider/department/employment-type/remote facet aggregates in a single call, a geo-radius nearby search, and a company/board search to see who's hiring and on which ATS. Filter by provider, department, location, employment type, remote, and open-role count; sort by relevance, posting date, open-role count, or company name. All billed pay-on-success.
The crawl updates continuously; this snapshot is dated July 21, 2026, and the most recently seen postings were observed the same day. Each board carries its own last-crawled timestamp and each posting its own first-seen/last-seen pair, so you can tell how current a given board or posting is.