Greenhouse, Lever & Ashby Jobs Scraper + Change Monitor

See which jobs appeared, changed or closed since your last check. Export current job listings and a downloadable changes CSV from Greenhouse, Lever (global/EU), and Ashby. Saved history makes the next run a comparison automatically.

Use it to keep a niche job feed current, track hiring at target companies, or review new roles before recruiting research. Updated events include the exact field values before and after an edit. Only public job postings are collected; no applicant data or ATS login is needed.

Current price, before 2026-10-19 04:40:54.985 UTC: $0.001 per successful complete board check, plus a $0.00005 start event at the default 256 MB. Current-job rows are included.

Scheduled price, from 2026-10-19 04:40:54.985 UTC: $0.0035 per successful complete board check + $0.00001 per current-job row exported to the default dataset + the unchanged $0.00005 start event. Every full export includes unchanged jobs in its row count. With writeCurrentJobs:false, no dataset-row charge applies. Failed/incomplete or budget-skipped boards have no board-check charge; the start event still applies. Check the Pricing tab for the active price used by your run.

Try one real board

1. Click Try for free / Start in Apify and keep the prefilled Brave board, or replace it with one company you follow.
2. Keep Remember previous results enabled and start the run. An Apify account and available credit are required; the Actor is pay per event.
3. Open Current jobs to inspect job titles, locations and links. Open Download changes CSV for a spreadsheet, or Change events for JSON.
4. Run the same input again later. New job IDs produce added, edited fields produce updated, and a missing job produces closed after two successful checks find it absent.

{
  "boards": ["https://boards.greenhouse.io/brave"],
  "persistState": true
}

The first run establishes a baseline: existing jobs appear as added. This does not mean they were posted today. A later unchanged run returns an empty changes list and a CSV with headers only; it still performs and charges for a successful board check.

Small real output sample

These two rows were collected from Brave on 2026-10-04 at 17:59 UTC during an owner test. This is a historical excerpt, not a promise that these roles are still open. It is not customer usage evidence.

Job	Location	Stable ID
Senior DevOps Engineer (browser CI/CD + AWS)	Anywhere	greenhouse:brave:1287724
Product Marketing Manager - Brave Search APIs	USA and EU	greenhouse:brave:7603697

What you receive

Output	Contents	Use it for
Current jobs (jobs)	Default dataset: title, location, job URL, description, source board, stableJobId, check time and provider-specific fields	JSON/CSV export; database upserts and job feeds
Download changes CSV (changesCsv)	CHANGES.csv: event type, job title/link, stable ID, changed field, before and after values	Reviewing changes in Excel or importing into a spreadsheet
Change events (changes)	EVENTS: added, updated, closed, with eventId and field-level before/after values	n8n/Make routing and downstream deduplication
Run summary (summary)	OUTPUT: per-board status, errors, events and carry-over snapshots	Checking whether each source completed successfully

CSV has one row per added/closed job and one row per changed field on updates. Arrays and objects are JSON within the cell. Spreadsheet formula-like text is prefixed with an apostrophe; the JSON event retains the original value. The CSV is written at the end of a normally completed run; use OUTPUT/EVENTS to inspect partial results after an interruption.

Greenhouse fields and meaningful changes

When present in the board-list response, Greenhouse first_published, requisition_id, metadata, and application_deadline are preserved as firstPublishedAt, requisitionId, metadata, and applicationDeadline. Missing fields stay absent; the Actor does not make extra per-job detail requests or infer deadlines. The listing API does not guarantee every detail field for every board.

Greenhouse description transport escaping is decoded once; already-HTML content retains its inner entities. Continue sanitizing employer HTML before displaying it. updatedAt is retained in job records but timestamp-only changes do not produce updated events. Title, description, location, deadlines and other content changes still do. Existing saved snapshots are migrated on the next successful check without treating newly available fields or HTML decoding alone as an employer edit.

Lower-write change monitoring (optional)

Set Export all current jobs (writeCurrentJobs) to false if you only consume changes. This skips current-job dataset writes, while EVENTS, CHANGES.csv, OUTPUT, comparison and saved history continue. OUTPUT.datasetMode reports omitted; the current-jobs dataset is empty in this mode. Default true preserves the full dataset contract. All boards are still fetched and checked completely; customer board-check prices are unchanged. This reduces dataset-write costs, not all platform costs or the price charged to the user.

{
  "boards": ["https://boards.greenhouse.io/brave"],
  "persistState": true,
  "monitorId": "daily-hiring-watch",
  "writeCurrentJobs": false
}

Supported boards

Paste 1–10 hosted company board URLs in one run. You supply the companies; this Actor does not discover all companies or search the entire web.

Provider	Supported input
Greenhouse	https://boards.greenhouse.io/brave or a job-boards.greenhouse.io company URL
Lever	Your company's https://jobs.lever.co/ board URL
Lever EU	Your company's https://jobs.eu.lever.co/ board URL
Ashby	Your company's https://jobs.ashbyhq.com/ board URL

Hosted job-detail URLs are also accepted. Custom company career domains, Workday, LinkedIn and Indeed are not supported. Lever pages are fully collected before comparison. Failed or incomplete responses preserve saved history and cannot close jobs.

Repeat daily with the same history

Save an Apify Task with the input below, substituting your own board URLs. Keep the same monitor name and board set on every run. For a different board set, create a different monitor name.

{
  "boards": ["https://boards.greenhouse.io/brave"],
  "persistState": true,
  "monitorId": "daily-hiring-watch"
}

Create a Schedule for that Task: daily at 09:00 UTC, cron 0 9 * * *, timezone Etc/UTC. Run each monitor sequentially and avoid overlapping manual runs. This is a configuration example; using this Actor does not create a schedule automatically.

For n8n or Make: trigger your workflow on Apify run success, read that run's EVENTS, route on type, and deduplicate by eventId. Upsert current jobs by stableJobId. Check OUTPUT.results as a successful run may still contain failed boards. Delivery to Slack, email or another destination belongs in your workflow; the Actor itself sends no messages.

Spending limits and complete output

The Actor checks the remaining budget before fetching a new board, then checks the exact cost of its complete export before writing any rows. It uses the run's actual pricing, including a dataset-row event if one is configured; the Pricing tab determines the active price used by that run, including during the announced notice period. A board that does not fit is skipped with reason: "spending-limit": no rows, new events or history advancement for that board. Earlier complete boards remain available. No board is silently shortened to fit a budget.

Check OUTPUT.complete as well as the run status. It is true only when every requested board has a complete result; a normally finished run can contain skipped or failed boards. Same-run resurrection preserves committed events and suppresses repeated row exports and board charges. If a dataset write has an uncertain outcome before a checkpoint is committed, the same run stops instead of retrying that export. Start a new run and inspect the failed run's partial dataset separately.

Large changes are checked before export: 20 MiB for OUTPUT, 10 MiB each for EVENTS and CHANGES.csv, and a conservative 256 MiB checkpoint-write allowance per execution attempt. The existing 10 MiB snapshot limit remains. When an output would exceed these limits, that whole board and the remaining new boards are skipped with reason: "output-limit"; existing history is retained. Split large batches. A single board whose complete output exceeds a limit cannot be processed by reducing the batch alone. There is no silent field, description, change or row truncation. These limits can reject large workloads accepted by older releases.

Per-board history checkpoints remain durable before board charging. Identical KV records and the redundant final checkpoint are not rewritten; a small export guard is stored before each dataset export. Storage, charging and dataset writes are still not one atomic transaction, so a failure can leave complete work uncharged or an explicitly failed run with partial rows.

Pricing and limits

Before 2026-10-19 04:40:54.985 UTC, each initial or unchanged successful board check remains $0.001, with job rows included. Ten boards once daily for 30 days at this rate cost $0.3015 in event charges, including 30 starts.

From that effective time, the approved scheduled prices are:

Event	Price	When it applies
Successful board check (board-sync)	$0.0035	Each complete baseline or repeat check, including unchanged or empty boards
Exported current job (apify-default-dataset-item)	$0.00001	Every current-job row written to the default dataset; none in omitted-dataset mode
Actor start (apify-actor-start)	$0.00005	Once at the default 256 MB

After the change, one board with 100 exported jobs costs $0.00455, and one with 1,000 exported jobs costs $0.01355, including one start. One board in omitted-dataset mode costs $0.00355, including the start. Change-event JSON, changes CSV and comparison history have no separate event price. Ten boards with 100 jobs each, once daily for 30 days entirely at the new rate, cost $1.3515 in event charges, assuming every check succeeds.

The start event is charged once up to 1 GB, then once per additional GB. Platform usage during the run is included in the event pricing; post-run storage and API use follow Apify's platform pricing. The scheduled increase preserves a notice period of at least 14 days; existing pricing remains active until the stated effective time. The run's pricing and the Pricing tab are authoritative.

Limits: 10 boards/run, 1,000 jobs/board, 8 MB/response and 10 MB carry-over state. Default memory is 256 MB, maximum 512 MB. No paid proxy or external LLM is used.

Saved history uses an account-owned named key-value store. Storage retention, deleted history, or a changed board set can require a new baseline. previousSnapshot can seed a new monitor; saved history takes precedence. Set persistState:false only when managing snapshots yourself.

Persisted observation IDs suppress replay charges, but saving and charging are not atomic: an interruption may leave completed work uncharged. Exactly-once delivery/billing and concurrent monitor writes are not guaranteed. Replay protection retains the latest 100 observation IDs; do not resurrect old runs after newer observations. Sanitize employer-provided description HTML before displaying it.