Company ATS Delta Snapshot Monitor

Monitor a known list of company job boards and receive compact hiring changes instead of repeatedly processing full job exports. The Actor reads public Greenhouse, Lever, and Ashby boards, normalizes the visible job fields, and compares each successful snapshot with the prior state in the same monitor namespace.

Use it for recurring account research, competitor hiring observation, or recruiting intelligence when you already know which companies matter. It does not discover arbitrary career sites, enrich people or companies, access applicant data, or infer why a job disappeared.

Quick start

{
  "monitorNamespace": "competitor-watchlist",
  "targets": [
    {
      "companyKey": "acme",
      "ats": "greenhouse",
      "boardToken": "acme"
    }
  ],
  "includeUnchanged": false
}

Keep monitorNamespace stable across scheduled runs. It selects the named key-value store that holds history. Use a different namespace for an independent watchlist or experiment.

Snapshot and delta semantics

The first successful run establishes a baseline. Because there is no earlier state, every current job appears in opened. Later runs compare jobs by the ATS job ID:

• opened: an ID exists now but not in the prior snapshot;
• closed: an ID existed previously but is absent now;
• changed: the same ID has a different title, department/team, location, employment type, or public job URL;
• unchanged: the tracked fields are identical.

changed does not currently mean that the full job description or compensation changed. A removed posting may have been filled, withdrawn, moved, or expired; the output does not claim which explanation is true.

Each successful dataset row represents one company snapshot. Failed targets produce an uncharged diagnostic row and do not stop healthy targets. If a previously non-empty board suddenly returns zero jobs, the Actor preserves the last good state and reports skipped_suspicious_empty. Set allowEmptySnapshots to true only after independently verifying that the board is genuinely empty.

Output

{
  "status": "succeeded",
  "companyKey": "acme",
  "ats": "greenhouse",
  "snapshotId": "2026-09-10T12:00:00.000Z",
  "billingEvent": "snapshot",
  "summary": {
    "currentJobs": 12,
    "opened": 2,
    "closed": 1,
    "changed": 3,
    "unchanged": 6
  },
  "opened": [],
  "closed": [],
  "changed": []
}

Billing and scheduling

The intended paid unit is the custom snapshot event: one successfully delivered company-board observation. Failed, suspicious-empty, and charge-limit-skipped targets never invoke that event and do not replace stored state. Before release, the Apify pricing configuration must disable or set a zero price for the synthetic apify-default-dataset-item event so diagnostic rows cannot create an unintended charge. The exact public snapshot price is pending cloud cost verification and Store approval; no price should be inferred from this repository.

Create an Apify Task with a stable input and schedule that Task at the cadence you need. Scheduled execution, persistent state, billing events, and platform cost must pass the private cloud release smoke before public use.

Limits, privacy, and support boundary

• One run accepts 1-100 unique ATS + companyKey targets and processes them sequentially.
• Each source request has a 15-second timeout. Public ATS availability and schemas can change without notice.
• Only public job-board data is fetched. Normalized snapshots are stored in the Actor user's Apify account under the selected named key-value store; results are stored in that run's dataset.
• dryRunNoPersist computes output without replacing history, but a successfully delivered snapshot remains a billable product event once pricing is enabled.
• The Actor is an independent community tool and is not affiliated with Greenhouse, Lever, Ashby, or monitored companies.
• v1 support covers the documented hosted ATS endpoints and schema. Custom career pages, authentication, proxy bypass, alerts, CRM setup, and investment or employment decisions are outside scope.