Workday Jobs Finder

Every open job of a company that hires through Workday, in one table, including the jobs beyond 2,000 that many Workday sites hide behind their list cap, with an honest count of anything out of reach, pay ranges parsed from the posting text, a map and a one-page hiring report. Run it once, or let it tell you only what is new each week. $1.50 per 1,000 jobs on the Free plan.

Quick start (no code)

1. Type the company name (Roche, NVIDIA, CVS Health) into Workday careers URLs or company names, one per line, or paste the careers URL if you have it (any page of the site works, even a single job's link). Click Start. A site with 2,000 jobs takes about two minutes.
2. Names are resolved against a bundled directory of public Workday career sites and each candidate is verified live before reading. The SUMMARY shows resolvedFrom, matchedBy and alternatives (other live sites of the same tenant, with their job counts), so a wrong pick is visible and fixable by pasting the alternative's URL. Names not in the directory come back as not_found with a hint.
3. On the run page: the Output tab is the table (tabs Jobs, Details, Locations, All fields; download as Excel, CSV or JSON). The Storage → Key-value store tab holds report.html (one-page hiring report, with a plain-words line per site: complete, incomplete and why, not usable), results-map.html (map) and SUMMARY (the same as JSON).
4. Is the list complete? In the SUMMARY: status: "found" and jobsReturned equal to jobsOnSite means yes. status: "partial" means Workday hid some jobs and the SUMMARY says how many and why.
5. Want it weekly? Tick Only new since my last run, then Schedule the Actor and pick a destination under Integrations.

The example tasks on this page are pre-filled; copy one and change the URL.

Which Actor for which company?

The company's jobs page URL contains …	Use
myworkdayjobs.com or myworkdaysite.com	this Actor (paste the URL, or just type the company name: it is looked up in a bundled directory of public Workday sites and verified live)
greenhouse.io, lever.co, ashbyhq.com, smartrecruiters.com, workable.com, recruitee.com, personio.de	ATS Jobs Finder (names are enough there)
something else	not covered yet

Known Workday sites to try: NVIDIA https://nvidia.wd5.myworkdayjobs.com/NVIDIAExternalCareerSite, Adobe https://adobe.wd5.myworkdayjobs.com/external_experienced, Salesforce https://salesforce.wd12.myworkdayjobs.com/External_Career_Site, Roche https://roche.wd3.myworkdayjobs.com/roche-ext, Novartis https://novartis.wd3.myworkdayjobs.com/Novartis_Careers, HP https://hp.wd5.myworkdayjobs.com/ExternalCareerSite, Target https://target.wd5.myworkdayjobs.com/targetcareers, CVS Health https://cvshealth.wd1.myworkdayjobs.com/CVS_Health_Careers, University of Pennsylvania https://wd1.myworkdaysite.com/recruiting/upenn/careers-at-penn.

Fields

Field	Notes
company, tenant, site, input	Workday tenant (e.g. nvidia) and site; the URL you gave
jobId, title, url, externalPath	always filled; jobId is the requisition id (e.g. JR2019657)
location, timeType, postedText, postedAt	from the list; postedAt is derived from postedText ("Posted 3 Days Ago"), exact up to 29 days, null for "30+ Days Ago"
startDate, country, remoteType, additionalLocations, hiringOrganization, applyUrl, description	only with Include description and exact date (one extra request per job)
salaryMin, salaryMax, salaryCurrency, salaryPeriod, salaryText, salaryRanges, salarySource	only with Include description and exact date or Only jobs that state pay: the pay range stated in the posting text (US and Canadian pay-transparency postings, e.g. 168,000 USD – 264,500 USD), parsed, with the exact quote in salaryText; several level tiers in one posting are merged to the lowest minimum and highest maximum with salaryRanges = how many. salarySource is always text; null when the posting states no range. Measured 2 Nov 2026: 36 of 36 US NVIDIA postings parsed, 0 invented values on the blind sample of 60 postings without pay
firstSeen	delta mode: the date you first received this posting
scrapedAt	when this run read the site

Workday has no structured salary field; what you get is what the posting says, nothing is estimated.

Example SUMMARY entry (key-value store):

{"input": "https://nvidia.wd5.myworkdayjobs.com/NVIDIAExternalCareerSite", "status": "found", "tenant": "nvidia", "site": "NVIDIAExternalCareerSite",
 "jobsOnSite": 2679, "listedTotal": 2679, "jobsMatched": 2679, "jobsReturned": 2679,
 "newJobs": 31, "closedJobs": 18, "unchangedJobs": 2648, "closed": [{"jobId": "JR…", "title": "…", "url": "…", "firstSeen": "2026-10-20", "closedOn": "2026-11-02"}],
 "snapshot": {"jobs": 2679, "remoteShare": 0.01, "newestPostedAt": "2026-11-02", "postedLast7Days": 190, "countriesTop": [["United States", 1200], ["India", 380]]}}

status is one of found, partial (error says how many jobs were out of reach and why), not_found (wrong tenant or site, or a name that resolved to nothing), robots_disallowed (the company's robots.txt forbids crawling the site; not read, not charged), not_workday_url, duplicate_of, error (Workday did not answer), stopped_cost_limit / skipped_cost_limit. A name input also carries resolvedFrom, matchedBy and alternatives. Every non-found entry carries a hint. Counts: listShows is the number Workday's own list displays (2,000 when capped), listedTotal the site's real size (from Workday's facets when the list is capped; listCapped: true then), jobsOnSite what was read, jobsMatched after your filters, jobsReturned what you received. Complete = status: "found" and jobsReturned equal to jobsOnSite equal to listedTotal.

Input

• Which Workday careers sites? URLs, one per line, with or without language prefix or job path. The same site given twice is read once.
• Narrow down (optional): Search text (Workday's own search, applied before reading, fast), Title keywords, Location keywords, Remote only, Only jobs that state pay, Posted within days, Max jobs per site (applied after the filters).
• Monitor: Only new since my last run and List name.
• Output: Include description and exact date, Map view.

Without Include description and exact date: postedAt is empty for jobs older than 29 days, jobs listed as "2 Locations" are not placed on the map and do not match location keywords, and Remote only can only read the location text. Turn it on when those matter; it costs one request per job (about one per second).

Monitor: only new since my last run

One mechanism, one switch. Tick Only new since my last run, schedule the Actor, choose a destination under Integrations. From the second run on you receive only postings that were not in your previous run, and the SUMMARY shows newJobs, unchangedJobs, closedJobs and the list closed. Only new postings are charged.

• The memory is a key-value store in your account named workday-jobs-finder-state-<list name>. Delete it to start over.
• The first run returns everything (firstRun: true).
• In delta mode the whole site is read each time (needed to detect closed postings); the cost cap limits only what you receive.
• Posted within days is a date filter without memory.

Why it is complete where others stop

Many Workday sites stop listing at 2,000 jobs and wrap around (NVIDIA: the list says 2,000, the facets 2,662); others (CVS Health, 18,791) page to the end. This Actor detects the cap, splits the site by Workday's own facets (nested ones included), merges the slices without duplicates and checks the count against the facets; uncapped sites are read to the end; a slice that cannot be split is reported as partial with the number out of reach. Measured live on 2 Nov 2026: Target 14,186 of 14,186, NVIDIA 2,679 of 2,679, CVS 18,762 of 18,786 during a 97-minute read while the site grew. Dates are requested in English, so postedText is always parseable.

Hiring report and map

Every run writes report.html (KPIs, jobs per city and country, postings per day over 28 days; printable) and, with Map view on, results-map.html (jobs by city; places resolved locally against GeoNames, CC BY 4.0). Both are in the run's key-value store.

Pricing and duration

Pay per job returned; sites not found and filtered-out jobs are free; platform usage is included. Per 1,000 jobs by Apify plan: Free and Bronze $1.50 · Silver $1.35 · Gold, Platinum and Diamond $1.15.

Run	Jobs	Cost on Free	Duration
Adobe, full list	~280	$0.42	15 s
NVIDIA, past the cap	~2,680	$4.02	~2.5 min
Target, past the cap	~14,200	$21.30	~12 min
10 sites, weekly monitor	~600 new per week	$0.90 per week	a few minutes
any site with Include description and exact date	n	same price	+ about 1 s per job

A cost cap: Console Run options → Maximum cost per run; API run parameter maxTotalChargeUsd. Without client-side filters the Actor reads only as many jobs as the cap pays for.

For developers and AI agents

from apify_client import ApifyClient
client = ApifyClient("<YOUR_API_TOKEN>")
run = client.actor("kohnschnee/workday-jobs-finder").call(
    run_input={"careerSites": ["https://nvidia.wd5.myworkdayjobs.com/NVIDIAExternalCareerSite"], "postedWithinDays": 7},
    max_total_charge_usd=1.0,
)
jobs = list(client.dataset(run["defaultDatasetId"]).iterate_items())
store = client.key_value_store(run["defaultKeyValueStoreId"])
summary = store.get_record("SUMMARY")["value"]
report_html = store.get_record("report.html")["value"]

n8n / Make / Zapier: the Apify node runs the Actor and hands over the dataset; fetch SUMMARY, report.html, results-map.html from the key-value store. AI agents: callable through Apify's MCP server. Key-value store keys: SUMMARY, SNAPSHOT, report.html, results-map.html. Dataset views: overview, salary (titled Details), places, full.

Limits, honestly

• One list request per second per host (two sites of one company share it), detail requests on a second lane; up to six sites are read in parallel. 2,000 jobs take about 100 seconds; 40 sites of 500 jobs each take about 3 minutes. Title keywords, Location keywords, Remote only and Only jobs that state pay are applied after reading; only Search text narrows what Workday sends. Posted within days stops the read early: Workday lists newest first, so paging ends once a whole page is older than asked (Adobe, 7 days: 120 of 278 read, all 70 matching jobs found, none missed against a full read; readStoppedEarly: true in the SUMMARY). Not in delta mode, which must see the whole site.
• Detail requests are made only for jobs that can still match: title keywords and plain place names are decided from the list first; only "2 Locations"-style rows need the detail to decide.
• A site that changes while it is read can differ from its live count: a job or two normally, a few per thousand on very large sites (CVS Health: 24 of 18,786 over 97 minutes). listedTotal versus jobsOnSite shows it.
• Companies that put Workday behind their own domain are covered only if the myworkdayjobs.com URL is known.
• Pay is parsed from the posting text only; postings that state no range get null, and a range written in an unusual form may be missed (open an issue with the posting URL).

Legal and fair use

The Actor reads the JSON endpoints that Workday careers sites expose to every visitor's browser, at one request per second, identified, without login, cookies or personal data. Before reading a site it fetches the tenant's robots.txt: most tenants list their career sites under Allow:; a few (Lowe's, Nike, Thermo Fisher on 2 Nov 2026) put them under Disallow: for everyone, and those sites are not read, by name or by URL, and come back as robots_disallowed with nothing charged. Names are resolved from a bundled directory built from public web indexes (618 tenants on 2 Nov 2026, growing) and from the tenant's own robots.txt; each candidate is verified with one list request. Respect the posting companies' terms when you reuse their postings.

FAQ

My URL comes back not_workday_url. It is not on myworkdayjobs.com or myworkdaysite.com. Open the company's jobs page, click a job, copy that address.

The Workday site says 2,000 jobs, I got more. That is the cap; the facets carry the real counts and the Actor reads each slice.

Why partial? A slice stayed at 2,000 and no facet could split it; the SUMMARY says how many jobs were out of reach. Narrow with Search text.

Why is postedAt empty for some jobs? Workday only says "Posted 30+ Days Ago" for them. Turn on Include description and exact date.

How do I get only new jobs each week? Only new since my last run plus a schedule.

How do I get pay ranges? Tick Only jobs that state pay (or Include description and exact date). The range the posting states is parsed into salaryMin/salaryMax; salaryText holds the quote so you can check it. Postings without a stated range are null, never estimated.

Where are the report and the map? Run page → Storage → Key-value store → report.html, results-map.html.

Something is wrong? Open an issue on this page with the URL that failed.

Changelog

• 1.3 (3 Nov 2026): company names and domains accepted (bundled directory of public Workday sites, verified live, alternatives listed), pay ranges parsed from the posting text (salaryMin…salarySource), Only jobs that state pay, Posted within days stops the read early on the newest-first list, detail requests only for jobs that can still match, URL parser no longer fails on malformed input.
• 1.2 (2 Nov 2026): delta mode with closed-postings list, report.html, Remote only, input form in sections with plain-language help, hints in SUMMARY, README rebuilt after an independent usability review.
• 1.1 (2 Nov 2026): hiring snapshot, map, four dataset views, tiered pricing; fixes from an independent code review (nested facets, uncapped tenants read to the end, partial with the number out of reach, duplicate URLs read once, per-site error state, myworkdaysite.com, English dates pinned, title filter before detail requests, exact cost-cap accounting).
• 1.0 (2 Nov 2026): first release.

Tested

Offline tests on recorded real responses and synthetic tenants shaped like Target, CVS and UPenn (23 tests), a live check against nine companies' own counts and random job URLs before each build, an independent code review and an independent usability review.