如何將公司招聘頁面的所有職位整理為 JSON(Greenhouse、Lever、Ashby 及另外 7 個 ATS)
How to get every job from a company's careers page as JSON (Greenhouse, Lever, Ashby and 7 more)
作者示範如何直接讀取公司的 ATS 公開職位 feed,將招聘頁面的職位整理為 JSON,無須爬取 LinkedIn。文中提供 Greenhouse、Lever、Ashby 的 Python 範例,並説明 10 個 ATS 的 feed 在資料格式、日期、地點、薪酬、分頁及新職位追蹤方面各有差異。
If you are building a job board, a hiring-signal tool for sales, or a "tell me when Company X posts a data role" alert, you do not need to scrape LinkedIn or fight anti-bot walls. Most companies run their careers page on an applicant tracking system (ATS), and every major ATS publishes the open jobs through a public feed that the careers page itself reads.
Here is how to use those feeds directly, what breaks when you do it at scale, and a shortcut if you would rather not maintain ten parsers.
Step 1: find out which ATS the company uses
Open the company's careers page and click any job. The link usually gives it away:
| If the job link goes to | The ATS is | Company identifier |
|---|---|---|
boards.greenhouse.io/airbnb/... |
Greenhouse | airbnb |
jobs.lever.co/spotify/... |
Lever | spotify |
jobs.ashbyhq.com/ramp/... |
Ashby | ramp |
jobs.smartrecruiters.com/BoschGroup/... |
SmartRecruiters | BoschGroup |
apply.workable.com/huggingface/... |
Workable | huggingface |
If the job opens on the company's own domain, view the page source and search for greenhouse, lever, ashby, workable or smartrecruiters: embedded boards load a script that contains the identifier.
Step 2: call the feed
import requests
def greenhouse(company):
r = requests.get(f"https://boards-api.greenhouse.io/v1/boards/{company}/jobs", params={"content": "true"}, timeout=30)
r.raise_for_status()
return [{"title": j["title"], "location": j["location"]["name"], "url": j["absolute_url"],
"department": ", ".join(d["name"] for d in j.get("departments", []))} for j in r.json()["jobs"]]
def lever(company):
r = requests.get(f"https://api.lever.co/v0/postings/{company}", params={"mode": "json"}, timeout=30)
r.raise_for_status()
return [{"title": j["text"], "location": j["categories"].get("location"), "url": j["hostedUrl"],
"team": j["categories"].get("team")} for j in r.json()]
def ashby(company):
r = requests.get(f"https://api.ashbyhq.com/posting-api/job-board/{company}", params={"includeCompensation": "true"}, timeout=30)
r.raise_for_status()
return [{"title": j["title"], "location": j.get("location"), "url": j.get("jobUrl"),
"salary": (j.get("compensation") or {}).get("compensationTierSummary")} for j in r.json()["jobs"]]
print(len(greenhouse("airbnb")), "Airbnb jobs")
print(len(lever("spotify")), "Spotify jobs")
print(len(ashby("ramp")), "Ramp jobs")
That is enough for a weekend project. A November 2026 run returned about 200 jobs for Airbnb, 90 for Spotify and 180 for Ramp, each in under a second.
The feed URLs for all ten ATS (including Workable, Recruitee, Personio, Teamtailor, Gem and Breezy HR), with what each returns, are in this reference repo: ats-job-feeds.
Step 3: what breaks at scale
Once you follow a few hundred companies, the work moves from fetching to normalising:
- Ten formats. Greenhouse and Ashby return JSON objects, Lever a JSON array, Personio XML, Teamtailor RSS. Dates come as ISO strings, epoch milliseconds or RFC 822.
-
Locations and remote work. "Remote (US)", "Hybrid - London", a list of offices, or a separate
workplaceTypefield, depending on the ATS. - Salary. Ashby gives a structured range for about half of jobs; Lever only when the employer fills it in; Greenhouse not at all in the public feed. When pay is written in the description, you need text parsing or an LLM to get numbers out.
- Paging. SmartRecruiters returns 100 postings per page; large employers have thousands.
- Only new jobs. For alerts you have to remember which IDs you already sent, per company and per filter set.
-
Identifiers that are not the company name.
Ubisoft2on SmartRecruiters, case-sensitive slugs on Lever.
The shortcut
I got tired of maintaining this and packaged it as an Apify Actor: Company Careers Jobs API. You paste careers pages, board URLs or plain company names; it finds the ATS, reads the public feed and returns one format for ten ATS (Greenhouse, Lever, Ashby, SmartRecruiters, Workable, Recruitee, Personio, Teamtailor, Gem, Breezy HR). It has an "only new jobs" mode for scheduled runs and optional AI fields (visa sponsorship with the quote, normalised salary, seniority, skills).
from apify_client import ApifyClient
client = ApifyClient("<YOUR_APIFY_TOKEN>")
run = client.actor("sandy_cornfield/company-jobs-api").call(run_input={
"companies": ["https://www.figma.com/careers/", "lever:spotify", "ashby:ramp"],
"keywords": ["engineer"],
"onlyNew": True,
})
for job in client.dataset(run["defaultDatasetId"]).iterate_items():
print(job["companySlug"], "|", job["title"], "|", job["workplaceType"], "|", job["url"])
It costs $2 per 1,000 jobs, with no monthly fee. If you only need one ATS, there are single-ATS versions for Greenhouse, Lever and Ashby.
Disclosure: I built those Actors. The feeds in steps 1 and 2 are free to use on their own; be polite with them (cache, a few requests per second at most).
來源:DEV Community · dev.to