ATS Jobs Intelligence

Fresh jobs straight from company career boards — before they hit Indeed — plus a hiring-intelligence summary for every company.

Most job scrapers fight Indeed's and LinkedIn's anti-bot defenses and hand you stale, deduplicated rows. This actor goes to the source: the public JSON APIs behind companies' own Applicant Tracking Systems (Greenhouse, Ashby, Lever). The data is fresh, structured, and un-scraped — and on top of it you get a per-company intelligence layer that answers the question raw rows can't: is this company scaling, and where?

Built for recruiters (fresh roles first), sales/BD teams (hiring is a buying signal), and competitive intel (watch a rival's org grow in real time).

---

Supported ATS providers

Provider	Board spec	Where the slug comes from
Greenhouse	greenhouse:slug	boards.greenhouse.io/<slug>
Ashby	ashby:slug	jobs.ashbyhq.com/<slug>
Lever	lever:slug	jobs.lever.co/<slug>

All three are public, unauthenticated endpoints (verified live). No API keys required.

---

Input

{
  "boards": ["greenhouse:airbnb", "ashby:ramp", "lever:veeva"],
  "keywords": ["engineer", "product"],
  "department": "Engineering",
  "location": "Remote",
  "remoteOnly": false,
  "postedWithinDays": 30,
  "maxJobsPerCompany": 1000,
  "includeIntelligence": true
}

Field	Type	Notes
boards (required)	string[]	"provider:slug" entries
keywords	string[]	Title must contain at least one (case-insensitive)
department	string	Department contains (case-insensitive)
location	string	Location contains (case-insensitive)
remoteOnly	boolean	Only remote-flagged roles
postedWithinDays	integer	Only roles published within N days (0 = off)
maxJobsPerCompany	integer	Cap on emitted job rows per board
includeIntelligence	boolean	Emit the per-company summary row (default true)
mode	string	snapshot (default) or watchlist — see below
watchlistName	string	Names the baseline in watchlist mode (default default)

---

Output

The dataset contains two record types, distinguished by recordType (with matching dataset views):

recordType: "job" — one per matched role:

provider, company, id, title, department, team, location,
remote, employmentType, publishedAt, updatedAt, url, applyUrl

recordType: "company_intelligence" — one per board (unless disabled):

totalOpenRoles, newRoles7d, newRoles30d, hiringVelocity,
pctOpenedLast30d, remotePct, topDepartments, topLocations,
fastestGrowingDepartment, seniorityMix, buyingSignals, summary

hiringVelocity is one of Aggressive (≥25% of roles opened in 30d), Active (≥10%), Steady (<10%), Frozen (0 new), or Unknown (no dated roles). Intelligence is computed on the company's full role set, so it reflects real hiring posture regardless of your keyword/location filter.

Example summary:

Ramp: 125 open roles, hiring aggressive · 41 opened in the last 30d · fastest-growing team: Sales · 38% remote. Signals: Scaling Sales — 6 new roles in 30d; Building leadership — 2 new director/exec role(s) in 30d.

---

Watchlist mode — track changes over time

Set mode: "watchlist" to turn the actor from a one-time snapshot into a change feed. Instead of returning every current job each run, it remembers the last run and reports only what changed since — ideal for a scheduled run (daily/weekly) that alerts you when target companies start (or stop) hiring.

How it works:

1. First run on a given watchlistName + board just saves a baseline (a free baseline row — nothing is charged).
2. Every run after diffs the live board against that baseline and emits recordType: "change" rows:

changeType	Meaning	Billed?
new_role	A role that wasn't there last run	✅ as job-scraped
removed_role	A role that's been filled or pulled	Free
company_shift	Posture digest: velocity change, new buying signals, new/removed counts	✅ as company-intelligence
no_change	Nothing changed for that board this run	Free
baseline	First-run seed	Free

3. The baseline advances to the current state after each run, so you only ever see genuinely new changes.

Use watchlistName to keep separate histories (e.g. competitors, target-accounts). Set includeIntelligence: false to get only role-level changes without the posture digest. See the Changes dataset view for a clean table.

Who it's for: sales/BD teams treating hiring as a buying signal, recruiters watching competitors open roles, and investors/analysts tracking a company's org growth week over week.

Tip: schedule watchlist runs in Apify (Schedules) and wire the dataset to a webhook/Slack for automatic hiring alerts.

Pricing

Pay per event — you are charged only for the results you receive:

• job-scraped — one charge per job row returned.
• company-intelligence — one charge per company hiring-intelligence summary.

Boards that error (wrong slug, company migrated off that ATS) are never charged — error rows are free. The intelligence summary is priced separately from job rows because it's the differentiator versus commodity single-board scrapers: set includeIntelligence: false to skip it and pay for job rows only.

---

Notes & limits

• Provider field coverage differs (from each ATS's public API): Greenhouse omits team/employmentType; Ashby and Lever omit updatedAt. Missing fields are null, never guessed.
• Lever's createdAt is epoch-ms and is normalized to ISO-8601 like the others.
• A board that errors (wrong slug, company migrated off that ATS) emits a recordType: "error" row and does not stop the run.
• Only isListed Ashby roles are returned.