DEV Community: Hải Hà Phan Thanh

Build a job feed from Greenhouse, Lever and Ashby career pages (no scraping browser needed)

Hải Hà Phan Thanh — Fri, 23 Oct 2026 02:27:51 +0000

Most tech companies do not host their own job pages. Their careers page is a thin wrapper around an applicant tracking system (ATS) such as Greenhouse, Lever or Ashby, and those vendors expose public JSON endpoints for job boards. That means you can build a job feed, a hiring-signal tracker or a salary dataset without a headless browser.

Here is how to do it yourself, and how to do it for eight ATS vendors at once with an Apify actor I built.

The DIY route: three endpoints

import requests

# Greenhouse
gh = requests.get("https://boards-api.greenhouse.io/v1/boards/stripe/jobs").json()["jobs"]
print(len(gh), gh[0]["title"], gh[0]["location"]["name"], gh[0]["absolute_url"])

# Lever (EU accounts use api.eu.lever.co)
lv = requests.get("https://api.lever.co/v0/postings/spotify?mode=json").json()
print(len(lv), lv[0]["text"], lv[0]["categories"].get("location"))

# Ashby
ab = requests.get(
    "https://api.ashbyhq.com/posting-api/job-board/openai",
    params={"includeCompensation": "true"}, timeout=20,
).json().get("jobs", [])
print(len(ab), ab[0]["title"])

(These are the vendors' public job-board endpoints; field names can change, so check their docs.) The field names differ for every vendor, dates come in different formats, and Workable, Workday, SmartRecruiters, Recruitee and BambooHR each need their own approach. Workday in particular needs a tenant, a datacenter number and a site name.

To find a company's slug, open a job on their careers page and read the URL: boards.greenhouse.io/<slug>, jobs.lever.co/<slug>, jobs.ashbyhq.com/<slug>.

One schema for eight ATS

My ATS Jobs Scraper actor covers Greenhouse, Lever, Ashby, Workable, Workday, SmartRecruiters, Recruitee and BambooHR and normalises them. No-code path: open the actor, paste board URLs (any mix) into companies, set keywords and locations, click Start, export CSV.

Python path:

# pip install apify-client pandas
from apify_client import ApifyClient
import pandas as pd

client = ApifyClient("YOUR_APIFY_TOKEN")
run = client.actor("bachiembmt/ats-jobs-scraper").call(run_input={
    "companies": [
        "https://boards.greenhouse.io/stripe",
        "https://jobs.lever.co/spotify",
        "ashby:openai",           # source:slug form is accepted
        "https://nvidia.wd5.myworkdayjobs.com/NVIDIAExternalCareerSite",
    ],
    "keywords": ["engineer"],
    "locations": ["remote", "london"],
    "postedWithinDays": 30,
    "maxJobsPerCompany": 25,
    "includeDescription": False,
})
df = pd.DataFrame(client.dataset(run.default_dataset_id).iterate_items())
print(df[["company", "title", "location", "salaryMin", "salaryMax", "url"]].head(20))

Output fields: source, company, jobId, title, department, team, location, remote, employmentType, salaryMin, salaryMax, salaryCurrency, postedAt, updatedAt, url, applyUrl, descriptionText.

A daily "new jobs" digest

Schedule the actor in the Console with postedWithinDays: 1, then fetch the last run's dataset:

last = client.actor("bachiembmt/ats-jobs-scraper").last_run(status="SUCCEEDED")
for j in last.dataset().iterate_items():
    if j["remote"]:
        print(f'{j["company"]}: {j["title"]} -> {j["url"]}')

Send that to Slack, email or a Google Sheet from there.

Limits you should know

Salary is patchy. Only Ashby, Lever and Recruitee publish pay ranges, and only when the employer fills them in. Greenhouse, Workday, SmartRecruiters, BambooHR and Workable have no public salary field.
Workday dates are approximate. The list only says "Posted 3 Days Ago", and 30+ days is reported as 30. With includeDescription you get the exact start date.
BambooHR list pages have no date; postedAt is filled only with descriptions on.
includeDescription costs one extra request per job on Workable, Workday, SmartRecruiters and BambooHR, at about 1 request per second.
You need the company's board slug. The actor does not discover which companies exist, and it only reads companies that use these eight vendors.
Only public, published postings. Boards that do not exist are skipped with a warning.

Price

Pay per event, charged only for jobs that pass your filters. It is about $1 per 1,000 jobs at the time of writing (check the Store page; I may adjust it). Scanning 50 company boards with ~40 jobs each is 2,000 jobs, so about $2.

For a handful of companies on Greenhouse or Lever, the DIY snippets are free and fine. I built the actor for the mixed-vendor case where normalising eight formats is the real work. Disclosure: it's my project.

Try it: ATS Jobs Scraper

How to get App Store and Google Play reviews into a spreadsheet in 5 minutes

Hải Hà Phan Thanh — Thu, 22 Oct 2026 03:20:02 +0000

If you work on an app, or on a competitor's, you eventually want the raw reviews in a spreadsheet: to tag complaints, track a rating drop after a release, or feed them to a sentiment model. The App Store and Google Play make that annoying. Apple has no export for other people's apps, Google Play's console only shows your own reviews, and the two stores have different formats.

This post shows two ways to do it: a DIY version for each store, and a version using an Apify actor I built that returns both stores in one schema. I'll be upfront about limits at the end.

The DIY route

Apple publishes a public customer-reviews RSS feed. No key needed:

import requests

app_id, country = "310633997", "us"   # WhatsApp
url = f"https://itunes.apple.com/{country}/rss/customerreviews/page=1/id={app_id}/sortby=mostrecent/json"
entries = requests.get(url, timeout=20).json()["feed"].get("entry", [])
for e in entries:
    if "im:rating" not in e:   # sometimes the first entry is the app itself, not a review
        continue
    print(e["im:rating"]["label"], e["title"]["label"])

Each page has up to 50 reviews and the feed stops at 10 pages, so roughly 500 of the newest reviews per app and country. Some apps return an empty feed even though they have reviews, which is the most common "my script is broken" report.

Google Play has no official public reviews API for third parties. The community library google-play-scraper works well:

from google_play_scraper import reviews, Sort
result, _ = reviews("com.whatsapp", lang="en", country="us", sort=Sort.NEWEST, count=200)

It parses Google's internal format, so it can break when Google changes things. For one app and one country this is all you need. The pain starts when you want 30 apps across 5 countries, with throttling, retries and one consistent column layout.

The 5-minute route: no code

Open the App Store & Play Reviews actor and click Try for free.
Paste app URLs or package names into apps. iOS and Android can be mixed; the store is detected automatically.
Set countries (for example us, gb), maxReviewsPerApp, and sort to newest.
Click Start, then Export the dataset as CSV or Excel. Open it in Google Sheets.

The Python route

# pip install apify-client pandas
from apify_client import ApifyClient
import pandas as pd

client = ApifyClient("YOUR_APIFY_TOKEN")
run = client.actor("bachiembmt/app-store-play-reviews").call(run_input={
    "apps": [
        "https://apps.apple.com/us/app/whatsapp-messenger/id310633997",
        "com.whatsapp",
    ],
    "countries": ["us", "gb"],
    "maxReviewsPerApp": 200,   # per app AND per country
    "sort": "newest",
    "language": "en",
})
items = list(client.dataset(run.default_dataset_id).iterate_items())
df = pd.DataFrame(items)
df.to_csv("reviews.csv", index=False)
print(df.groupby(["appName", "store"])["rating"].agg(["count", "mean"]))

Every row has the same fields regardless of store: store, appId, appName, country, reviewId, userName, rating, title, text, version, date, thumbsUp, developerReply, replyDate, url.

Monitoring only new reviews

Use minDate with sort: newest and schedule the actor weekly in the Console:

run_input = {"apps": ["com.whatsapp"], "sort": "newest", "minDate": "2026-10-15"}

Then a quick check for what hurts most:

low = df[df.rating <= 2]
print(low.groupby("version").size().sort_values(ascending=False).head())
print(low.text.str.len().describe())

Pipe low.text into whatever sentiment or topic tool you prefer. The unified schema means you write that code once.

Honest limits

Apple: the free RSS feed stops at about 500 reviews per app and country. The actor reads Apple's web reviews API instead (several hundred deep per country in my tests) and falls back to RSS, but it is still not a full multi-year history.
Google Play access depends on an unofficial library, so it can break for a while after Google changes its format.
Requests are throttled to about 1 per second, so big jobs take a while.
Only public review data. No login, no private data.
If you only need one app once, the DIY snippets above are free and enough.

Price

Pay per event: about $0.50 per 1,000 reviews returned (check the Store page for the current price). 200 reviews cost around a tenth of a cent; a 30-app by 3-country run at 200 each is 18,000 reviews, so around $9. Apify accounts normally include some free monthly platform credit, so a first test costs nothing; check your plan for the current amount.

I built this actor, so I'm biased, but I'd rather you try the free snippets first and use the actor only if maintaining the scraper is not how you want to spend your time.

Try it: App Store & Play Reviews actor