Gaming Industry Jobs & Hiring Tracker

Collect public vacancies from the SmartRecruiters boards of Entain, Evolution and Playtech. Export jobs with employer, title, location, department, employment type and source links. Supply a previous snapshot to identify newly observed jobs, field changes and jobs no longer observed within the same filters.

Use the results for recruiting research, gaming industry vacancy feeds and changes in the mix of advertised roles. Coverage is limited to these three public boards; it does not include every vacancy at each employer. This is an independent tool, unaffiliated with the employers or SmartRecruiters.

First run

Select an employer and leave the filters empty. Optionally set Max total charge in run options to bound paid delivery; API clients use maxTotalChargeUsd. A verified unset limit permits delivery within the source and size bounds. An explicit zero prevents paid rows. Check the price shown in the Actor's pricing panel before starting.

Example input for one employer:

{
  "companies": ["Evolution"],
  "maxJobsPerCompany": 500,
  "countries": [],
  "titleKeywords": []
}

Omitting the input record or supplying {} selects all three employers, scans up to 500 source jobs per employer and applies no filters. The form starts with Evolution only for a smaller first run. There is no invented demo output: a normal run requests current public source data.

Input

Field	Meaning
companies	One to three of Entain, Evolution, Playtech. Defaults to all three.
maxJobsPerCompany	Source jobs scanned per employer, from 1 to 1000; default 500. This is not the number of filtered matches.
countries	Optional country codes such as GE, MT, GB. Up to 30 distinct codes. Empty selects all countries.
titleKeywords	Up to 10 words or phrases, each at most 100 characters. A title must contain at least one phrase, ignoring case. Empty selects all titles.
previousSnapshot	Optional JSON string containing the complete SNAPSHOT record from a previous run, up to 4 MiB of UTF-8 JSON. Leave empty for collection only.

Countries and title filters both apply when supplied. Filtering happens after source traversal, so a small scan limit can miss matching jobs later in the board. Employers are processed alphabetically under a shared 90-second source collection limit, with at most 20 seconds per request. A slow earlier board can leave later employers incomplete. Use a run timeout of at least 180 seconds to allow collection and delivery.

Results

The default Dataset contains rows with recordType: "job". Each row includes:

• company, employerName, jobId, jobKey, title.
• country, region, city, location, remote, hybrid.
• department, function, industry, employmentType, experienceLevel.
• releasedDate, jobUrl, sourceUrl, and billing status at the time of writing.

Optional values absent from the source are null. releasedDate is the source publication timestamp, not a last-edit timestamp. Remote/hybrid flags are preserved when provided. The tool does not infer salary, employees, hires or headcount, and does not collect applicants or advertisement body text.

Diagnostics share the Dataset with jobs and have recordType: "diagnostic". Filter by recordType when importing vacancies into a CRM or spreadsheet.

The default key-value store can contain:

Record	Contents
OUTPUT	Outcome, per-employer coverage, observed dates, delivered counts and vacancy mix. Check this first.
SNAPSHOT	Versioned source snapshot for your next comparison. Available only after all collected job rows have been delivered and their result events confirmed.
COMPARISON	Detailed changes, present only when a previous snapshot was supplied and full delivery completed. Inspect withheldCompanies before using changes.

completed, partial and budget_stopped normally finish with a SUCCEEDED run status; source_unavailable finishes with FAILED. A successful platform status alone does not mean complete source coverage. A cap, source error, changing total, duplicate, rejected row or unverified empty board makes that employer incomplete. Valid rows from incomplete boards remain usable and billable. Complete traversal of a changing board is not a transactional snapshot.

Common coverage reasons: SOURCE_LIMIT means the scan cap was reached; SOURCE_DEADLINE means the shared collection time expired; EMPTY_BOARD_UNVERIFIED means an empty board could not be confirmed; SOURCE_TOTAL_CHANGED means source counts changed during traversal; DUPLICATE_POSTING_ID means repeated source IDs were found. INCOMPLETE_TRAVERSAL means the returned pages did not account for the reported board total. These are coverage limitations, not proof that missing vacancies closed.

Compare two observations

Download SNAPSHOT after the first run. On a later run, use the same employers, countries and title keywords, and pass the downloaded object as a JSON-encoded string in previousSnapshot:

const input = {
  companies: previous.companies,
  countries: previous.filters.countries,
  titleKeywords: previous.filters.titleKeywords,
  maxJobsPerCompany: 500,
  previousSnapshot: JSON.stringify(previous)
};

Here previous is the parsed JSON downloaded from the first run. Observation windows must not overlap. Comparisons are withheld for employers whose earlier or current collection is incomplete. A changed filter scope is rejected.

Changes are newly_observed, changed and no_longer_observed. A job no longer observed may have left the filters or disappeared from the board; that does not establish a hire or closure. Field changes compare literal values, including differently formatted timestamps. Snapshot digests detect inconsistent data; they are not an authenticity signature.

Pricing and recovery

One result event is requested for each delivered vacancy row. Repeat observations deliver and charge current vacancies again, including unchanged ones. Comparison artifacts do not request a separate result event. Diagnostics and zero matching jobs request no vacancy result events; any platform Actor-start charge displayed in the pricing panel can still apply.

Reserve enough of your Max total charge for any start event plus the expected vacancy rows. If the remaining limit cannot cover another job, delivery stops and full snapshot/comparison artifacts are withheld. Already delivered rows remain in the Dataset. A row's unconfirmed_at_write billing annotation is not a final payment receipt. Check OUTPUT when present and the run's settled event counters. On a failed run, OUTPUT can be absent even after rows have been delivered or charged; inspect the Dataset, run status message and settled counters directly.

Automatic retry, resurrection and migration recovery of the same run are unsupported. A durable run claim is saved before collection; a restarted container finding that claim stops before collecting or delivering again. This prevents automatic replay but does not complete interrupted work. After an uncertain storage or charging response, inspect its Dataset, artifacts and charges before starting a new run; a new run can deliver and charge overlapping jobs again. Full snapshots are limited to 4 MiB; comparison exports to 9 MiB. These are product limits.

Sources and support

Data comes from the SmartRecruiters public Posting API for the three named employers. Source totals and observation dates are included so you can assess coverage.

If output is incomplete or a field appears incorrect, open an Actor issue with the run link, employer and expected behavior. Do not include credentials or applicant data.