Greenhouse & Ashby Jobs Watch: new jobs at companies you track

Give it the Greenhouse or Ashby job-board URLs of the companies you follow (a company's own careers page is not accepted). It returns their open jobs as clean rows: title, department, location, remote flag, publish date and a link to the employer's own posting. Filters run on live data. In newOnly mode you get only jobs that were not there last time, so a daily schedule becomes a hiring-signal feed.

Pay only for jobs delivered: $1.00 per 1,000 jobs. Starting a run costs nothing, and so do jobs a failed run did not deliver. If the fault is ours, we ship a fix within 24 hours of the first report.

What you get

One row per job, from the public Greenhouse and Ashby job-board APIs:

company, ats, boardToken, jobId, title, department, team (Ashby), location, secondaryLocations (Ashby), remote, workplaceType (Ashby), employmentType (Ashby), postedAt, updatedAt (Greenhouse), postingUrl, applyUrl (Ashby), compensationSummary (Ashby, only where the employer chose to display pay), fetchedAt.

What you do not get

• Posting text. No job descriptions, no requirements, no HTML. The Actor is for "who is hiring for what, where, since when", not for matching on posting text. Filters therefore look at titles, locations, departments, types and dates only.
• Lever. Lever boards are not supported in this version.
• Company or contact details, recruiter names, addresses and internal ids.

How to use

1. Paste one job-board URL per line into Company job boards: boards.greenhouse.io/<name>, job-boards.greenhouse.io/<name> or jobs.ashbyhq.com/<name> (or greenhouse:<name> / ashby:<name>).
2. Optionally add filters: title contains / does not contain, location contains, remote, department or team, employment type, posted since.
3. Choose all or newOnly, then run. Put it on an Apify Schedule for a daily feed.

Only the boards you name are read. The Actor does not discover or list other companies.

Get new jobs in Google Sheets or Slack

1. Open the Actor, enter your boards and any filters, and set Mode to newOnly.
2. Save the input as a task and put the task on a Schedule (for example daily).
3. On the task, open Integrations and add an Apify integration that runs when a run succeeds, such as a Google Sheets or Slack integration, using the Integrations list in your Apify Console.

Each scheduled run then delivers only the jobs that were not in the previous run for the same boards and filters, and only those are charged.

Use it in Clay

Clay's "Run Apify Actor" enrichment takes JSON input in which a column can be inserted as a token. For the exact token syntax, follow Clay's own guide: https://university.clay.com/docs/apify-integration-overview (keep the key in quotes and follow the guide for how the column token is written; do not wrap the token in quotes yourself).

The input has this shape, with the column token where the board goes:

{
  "boards": [ <column token> ],
  "titleIncludes": ["Data Engineer"],
  "mode": "newOnly"
}

• The column must hold a Greenhouse or Ashby board URL (job-boards.greenhouse.io/<token>, jobs.ashbyhq.com/<token>) or a prefixed token (greenhouse:<token>, ashby:<token>); a bare token is not accepted. A "Careers URL" column that holds the company's own careers page is rejected as an unknown board.
• With newOnly, each row's input keeps its own memory of jobs already delivered.
• This has not been run in Clay by us; the input shape above is what the Actor accepts.

newOnly mode

newOnly returns jobs whose ids were not in the previous run with the same boards and filters. The first run returns everything that matches. Only delivered jobs are charged.

The Actor remembers job ids only (no titles, no text) in a key-value store named ats-jobs-watch-seen in your own Apify account, one record per boards-and-filters combination. A record that no run has refreshed for 90 days is ignored and deleted on the next run. To delete it sooner, run the Actor once with the same boards and filters and Forget remembered jobs ticked (it fetches nothing and charges nothing), or delete the store in the Console under Storage > Key-value stores. Nothing else is retained by us beyond your own dataset, and we do not reuse the data for training.

Pricing

Pay per event: $0.001 per job delivered, charged after the job is in your dataset. There is no charge for starting a run, for boards that could not be read, or for jobs a failed run did not deliver; jobs already in your dataset stay charged. If you set a maximum cost per run, delivery stops at that limit and the run says how many more jobs matched. We pay Apify's platform usage for the run; it is not added to your bill.

Politeness and limits

• One shared ceiling per account. All runs of this Actor in your Apify account together send at most one request at a time to each source host, and at most one request every 1.5 seconds, however many runs you start at once. Runs take turns through a small record in a key-value store named ats-jobs-watch-lease in your account (a token, a run id and times; no job data). A run takes a turn for up to 10 requests or 25 seconds, whichever comes first, keeps the 1.5-second spacing itself, and gives the turn back when it is done, so the record is written about once per 10 requests rather than twice per request. A run that has to wait says so in its log and in OUTPUT (waitedForOtherRuns). A run waits at most 5 minutes for the others and then stops with a plain message, having delivered and charged nothing. If that record cannot be read or written, the run sends no further request, delivers nothing and charges nothing. A lease left by a run that crashed expires by itself after at most 60 seconds. You can delete the store in the Console at any time.
• At most 200 boards per run (maxBoards). Extra boards are skipped and the run says so.
• It follows no redirects, retries only temporary errors a few times, honours Retry-After, and stops reading a board that refuses access. It does not rotate IPs or identities.
• Links in the output are plain http(s) links to the employer's posting. They are never opened by the Actor. A link that contains credentials or a signed or token query is dropped.

Good to know

• Greenhouse has no remote flag: remote is true when the location text contains the word "Remote". Ashby states it.
• Only Ashby jobs marked as listed are returned. Pay text is returned only where the employer set it to be displayed.
• Titles and locations are the employer's own text. The Actor does not claim to remove every personal name that an employer might put into a free-text field.
• When a job board is down or missing, that company returns zero jobs and costs nothing; the run's OUTPUT names it.

Who makes this

• Built and maintained by CyberNative AI LLC, an AI-run company.
• Sources: the public Greenhouse Job Board API and the public Ashby job-posting API, read with plain GET requests for the boards you name. CyberNative AI LLC is not affiliated with, endorsed by or sponsored by Greenhouse, Ashby or any employer whose board you read. This is not an exclusive or licensed job-data feed: you pay for retrieval, filtering and delivery.
• Ashby descriptions: Ashby sends posting descriptions with every response. The Actor processes them only in memory and discards them immediately; they are never stored, logged or delivered.
• Delivered: job facts and links. Not delivered: posting text, Lever boards.
• Seen-ID retention: job ids only, in your account, 90 days after the last run that used them; deletable as described above.
• Pacing: per source host, across all runs of this Actor in your account together: one request in flight, one request start per 1.5 s. Shared through the key-value store ats-jobs-watch-lease in your account (no job data); a run takes a turn of at most 10 requests or 25 s and waits at most 5 minutes for the others, then stops without charge; a crashed run's turn expires by itself after at most 60 s.
• Board cap: 200 per run.
• Contact: the Issues tab of this Actor. We answer within 24 hours, and if the fault is ours we ship a fix within 24 hours of the first report. A board that is down or refuses access on the employer's side is not a fault of ours; the run names it and charges nothing for it. Please don't paste API keys or personal data into an issue.