Files
jobboard/docs/plan-hosts/own-2.md
T

5.9 KiB

Own-site career pages, batch 2 (30 companies)

Every request was a live curl on 2026-09-29.

Counts per class

class n companies
known-host 12 mews, cognigy, solaris, airtame, automattic, metacore, rovio, vivino, manomano, dexter-energy, glovo, schibsted
new-host 10 zeiss, danske-bank, seaborg, depop, onfido, snyk, tink, allegro, einride, klarna
own-api 3 maersk, bending-spoons, amazon
jsonld 1 odoo
browser 3 adjust, personio, tradeshift
none 1 wetransfer

Counts per host (primary host of each company)

  • greenhouse.io: 6 (mews mewssystems, cognigy nice filtered, solaris solarisbank, airtame airtamejobs, automattic automatticcareers, metacore metacore)
  • teamtailor: 3 (vivino, manomano, schibsted; vivino and manomano on custom domains)
  • lever.co: 1 (rovio rovio-2); smartrecruiters.com: 1 (glovo, via the DeliveryHero board); recruitee.com: 1 (dexter-energy dexterenergyservices)
  • myworkdayjobs.com: 5 (zeiss, depop, onfido, snyk, tink); Maersk also uses Workday and SAP behind its own API
  • oraclecloud.com: 1 (danske-bank); hr-manager.net: 1 (seaborg); jobylon.com: 1 (einride; schibsted has a second Jobylon feed); successfactors: 1 (allegro); jobs.deel.com: 1 (klarna)
  • own site: odoo, maersk, bending-spoons, amazon, wetransfer (via parent)
  • not readable by curl: adjust and personio (Vercel checkpoint), tradeshift (list source unknown)

How each new-host listing is read

Workday (myworkdayjobs.com). POST JSON to https://<tenant>.<wdN>.myworkdayjobs.com/wday/cxs/<tenant>/<site>/jobs with {"appliedFacets":{},"limit":20,"offset":0,"searchText":""}. The response has total and jobPostings with title, externalPath, locationsText; no login or cookie was needed. Verified for zeiss (zeissgroup.wd3, site External, 847), depop (etsy.wd5, site Depop_Careers, 20), onfido (entrust.wd1, site EntrustCareers, 81), snyk (snyk.wd103, site External, 21) and tink (visa.wd5, site Visa, 760). Onfido and Tink are inside their acquirer's board and can only be separated by searchText or location facets. The tenant of zeiss, onfido and depop was found through web search, not from the company page (zeiss and onfido pages are a JS widget and a Cloudflare challenge respectively).

Oracle Recruiting Cloud (danske-bank). Public REST without login: GET https://ejqi.fa.ocs.oraclecloud.eu/hcmRestApi/resources/latest/recruitingCEJobRequisitions?onlyData=true&finder=findReqs;siteNumber=CX_1001,limit=25,offset=0 returns items[0].TotalJobsCount (135) and items[0].requisitionList. Job pages are https://ejqi.fa.ocs.oraclecloud.eu/hcmUI/CandidateExperience/en/sites/CX_1001/job/<id>. Only limit=1 was requested, so paging with larger limits is not tested.

HR Manager / Talentech (seaborg). GET https://api.hr-manager.net/JobPortal.svc/<customer>/PositionList/rss/?incads=true returns an RSS channel; customer is seaborg_tr. It returned a valid channel with zero items, matching the portal's "no positions" text. Item format was therefore not seen. Denmark and Norway use this host for many companies, so the reader is probably reusable.

SAP SuccessFactors (allegro). Career site on a custom domain (careers.allegro.eu, tenant career55.sapsf.eu, company allegrospz). GET https://careers.allegro.eu/sitemap.xml lists every job as https://careers.allegro.eu/job/<slug>/<id>/ (312 seen), and a job page carries JSON-LD JobPosting. The candidate-facing career55.sapsf.eu/career?... page is a sign-in shell. The SmartRecruiters company Allegro remains but only holds a notice that the company moved to SAP.

Jobylon (einride). No JSON found. The widget iframe GET https://cdn.jobylon.com/jobs/companies/<company_id>/embed/v2/?target=jobylon-jobs-widget&page_size=50 (company id 3143) is HTML that assigns JBL.embed_v2['jobs'] = [ {id: '...', url: '/jobs/<id>-<slug>/', ...}, ...] (33 entries), which can be parsed. Jobylon also has per-company feeds of the form https://feed.jobylon.com/feeds/<uuid>/?format=json, but the uuid is set by the company; Schibsted's public feed (41f7d609-c6ba-4fc0-9032-dcb01f8db52d, 4 postings) is in its theme bundle and works without login. No feed uuid was found for Einride.

Deel job board (klarna). https://jobs.deel.com/<slug> (slug klarna) is a Next.js page whose server payload contains "jobPostings":[{"id":...,"jobId":...,"title":...}] (95 entries); job pages https://jobs.deel.com/klarna/job-details/<uuid>/overview carry JSON-LD JobPosting. No public JSON API was found. Whether the list is paginated beyond the 95 was not tested.

Things to know before building

  • Slugs that differ from the company name (naive slug guessing fails): mews -> mewssystems, solaris -> solarisbank, airtame -> airtamejobs, automattic -> automatticcareers, rovio -> rovio-2, dexter-energy -> dexterenergyservices, cognigy -> shared nice board with a metadata filter (field id 4952164101 = Cognigy).
  • Glovo roles are a subset of the Delivery Hero SmartRecruiters board (987 roles); the 270 Glovo-only roles were seen only on careers.glovoapp.com through a WordPress admin-ajax.php action.
  • Maersk's list API returns 401 unless a constant Consumer-Key header (taken from the site's JS) is sent; the app owner should decide whether that is acceptable.
  • Companies now inside another one: onfido (Entrust), tink (Visa), wetransfer (Bending Spoons), cognigy (NICE), adjust (AppLovin, unverified), seaborg (now Saltfoss).
  • automattic.com answered 403 to a Chrome-like User-Agent but 200 to Mozilla/5.0; other sites blocked curl outright: adjust.com and personio.com (Vercel Security Checkpoint, HTTP 429), allegro.pl (DataDome), entrust.com, manomano.fr and rovio.com feed (Cloudflare), depop.com (403).
  • The jobs_url for automattic is the Greenhouse board, because automattic.com/work-with-us/jobs/ currently shows only a scam warning and no roles.