Dataset

Every open role, from the company's own career system_

Job postings read straight from the applicant tracking systems behind company career sites, with job portals filling the gaps. Normalized by function, level and location, joined to the company record, and delivered as files on your schedule.

Jobs datasetWhat you getBulk
  • 10M+ open roles from 151k company career sites
  • Read from ~70 applicant tracking systems, refreshed hourly
  • Normalized function, level and location; tools named in the posting
  • CSV, JSONL or Parquet, to your bucket or a signed link
10M+open roles
151kcareer sites
~70ATS read directly
Hourlyingestion
00 — Sources

Read from the systems companies hire with

The applicant tracking system behind each company's career site first; job portals fill in companies that do not run one.

Applicant tracking systems~70 systems

Read straight from the career site a company runs.

Greenhouse
Lever
Ashby
Workable
SmartRecruiters
Personio
Recruitee
Teamtailor
Workday
BambooHR
JazzHR
Breezy HR
Jobvite
iCIMS
Pinpoint
JOIN
+ 50 more
Job portalsGap filling

For companies without their own career system.

LinkedIn
Indeed
Glassdoor
StepStone
Welcome to the Jungle
XING
Wellfound
Monster
01 — Schema

What is in the files

2 files, joined on the company domain. Pick one to see its fields and a sample record.

One row per posting, open or recently closed.

Identity
job_idStable id across refreshes
urlThe posting on the company's career site
atsThe applicant tracking system it came from
Posting
titleThe title as written
description_mdFull description, cleaned to Markdown
first_seen_atWhen we first saw it
last_seen_atWhen we last saw it open
closed_atWhen it disappeared, if it has
Classification
functionSales, marketing, engineering, support, finance…
levelC-level, VP, Head, Director, Manager, IC
employment_typeFull-time, part-time, contract, internship
Location
cityNormalized city
country_codeISO 3166 alpha-2
remoteRemote, hybrid or on-site
Signals
tools_mentionedSoftware named in the posting
languagesLanguages the role asks for
Company
companyNested company record: domain, name, size, country, offering
Sample record · illustrative
{
  "job_id": "j_8f2c41",
  "url": "https://jobs.example-ats.com/shipvane/sdr-berlin",
  "ats": "example-ats",
  "title": "Sales Development Representative (DACH)",
  "function": "sales",
  "level": "ic",
  "employment_type": "full_time",
  "city": "Berlin",
  "country_code": "DE",
  "remote": "hybrid",
  "tools_mentioned": ["hubspot", "outreach"],
  "first_seen_at": "2026-10-14T06:12:00Z",
  "last_seen_at": "2026-10-24T05:58:00Z",
  "company": { "domain": "shipvane.io", "employees": "51-200" }
}
02 — Delivery

On your schedule, in your stack

Formats

Parquet, CSV or JSONL

Parquet for warehouses, CSV for everything else, JSONL for pipelines and language models.

Destinations

Your bucket or a link

Files land in your S3 or GCS bucket, or arrive as a signed download link.

Refresh

Daily full snapshot

Hourly deltas on request. Weekly or monthly for research use.

01

Pick a scope

Countries, segments, functions, or a list of domains.

02

Check a sample

A sample file for your scope, before you commit.

03

We set up delivery

Format, destination and schedule.

04

Files land

On schedule, with a manifest per drop.

03 — Quality

Built so you do not have to clean it

From the source

Postings come from the company's own ATS first, so reposts collapse and agency listings stay out.

One posting, one row

The same role on several boards is still one record, with a stable id.

Normalized

Function, level and location are mapped to fixed values, so you filter instead of parsing titles.

Opened and closed

First seen, last seen and closed dates let you measure time to hire and hiring velocity.

Joined to the company

Every posting carries the company domain, size and what the company sells.

Clean descriptions

HTML stripped to Markdown, ready for search or a language model.

04 — Who uses it

Teams that build on this data

Sales tools

Hiring signals inside your product, without running a crawler.

Investors and research

Hiring velocity by sector, country and function.

Job boards

Backfill with roles straight from career sites.

Data teams

Hiring as a feature in scoring and territory models.

Hiring data

Not ready for bulk? Use it live

The same data as a search, over MCP and the API, as bulk files, or pushed to you as it changes.

Plays

Plays on the same data

05 — FAQ

Jobs dataset: questions

Where do the postings come from?

From about 70 applicant tracking systems that host company career sites, plus job portals for companies without one.

How fresh is it?

We check career sites every hour. A daily file contains everything open at the time of the snapshot plus what closed since the last one.

Can I get only some countries or functions?

Yes. A dataset can be scoped by country, function, company size or a list of domains.

Which formats and destinations?

CSV, JSONL or Parquet, delivered to your S3 or GCS bucket or as a signed download link.

Can I try it first?

Yes: ask for a sample file for your scope, or query the same data over the API with 1,000 free records.

More datasets

Other datasets

Get a sample of
the jobs data

Tell us your scope and we send a sample file. Or query the same data over the API with 1,000 free records.

  1. 01Sign up, no card
  2. 02Connect Claude, ChatGPT or an API key
  3. 03Ask for your market, 1,000 records free