Remote Job Feed with Geo-Eligibility

Remote jobs you can actually do from your country.

This Actor collects job postings from public Greenhouse, Lever and Ashby job boards and works out, for every job, which countries you may work from. "Remote" on a job board often means "remote, but US only" or "remote within EMEA". This Actor turns that into structured data you can filter on.

• 🌍 eligibleCountries, remoteScope (worldwide / regional / country) and isWorldwide on every job
• 🔎 Evidence for every decision: the exact location text or sentence it is based on
• 🎯 Filters: country, remote/hybrid/on-site, keywords, departments, posting date, employment and contract type
• 💸 You only pay for jobs that pass your filters. Jobs that are filtered out are free.
• 🧾 Built-in, verified list of 240+ companies with public boards, so you can start without collecting slugs

Who is this for?

• Job board operators who want a feed of remote jobs per country or region
• Recruiters and sourcers tracking hiring at specific companies
• Job seekers who are tired of clicking "Remote" jobs that turn out to be US-only

How to use it

1. Add companies: board slugs (gitlab), prefixed slugs (ashby:linear, lever:palantir, greenhouse:stripe) or board URLs (https://jobs.ashbyhq.com/linear). Or tick Add the built-in list of remote-friendly companies.
2. Choose the countries you want to work from, for example Germany.
3. Optionally narrow down: workplace type, keywords, departments, "posted within the last N days".
4. Run. Results appear in the dataset; a per-company report is saved as RUN_REPORT in the key-value store.

Example input

{
    "companies": ["gitlab", "ashby:linear", "https://job-boards.greenhouse.io/grafanalabs"],
    "useStarterList": false,
    "countries": ["DE"],
    "workplaceTypes": ["remote"],
    "keywords": ["engineer", "developer"],
    "postedWithinDays": 30
}

Example output (shortened)

{
    "id": "greenhouse:algolia:5854606004",
    "companyName": "Algolia",
    "title": "Commercial Account Executive, DACH",
    "url": "https://job-boards.greenhouse.io/algolia/jobs/5854606004",
    "locations": ["Remote - Germany"],
    "workplaceType": "remote",
    "remoteScope": "single-country",
    "remoteScopeLabel": "Germany only",
    "isWorldwide": false,
    "eligibleCountries": ["DE"],
    "eligibleRegions": [],
    "eligibilityConfidence": "high",
    "eligibilityEvidence": [{ "source": "location", "text": "Remote - Germany" }],
    "timezoneRequirement": null,
    "employmentType": "unknown",
    "engagementType": "unknown",
    "postedAt": "2026-04-07T15:53:42.000Z",
    "matchedCountries": ["DE"],
    "descriptionText": "Algolia is ..."
}

Output fields

Field	Description
id	Stable unique id: source:board:jobId
source, companySlug, companyName, jobId	Where the job comes from (Lever and Ashby publish no company name: it comes from the built-in list, otherwise from the board slug, e.g. ava-labs becomes Ava Labs)
title, url, applyUrl, department, team, locations	As published by the company
workplaceType	remote, hybrid, onsite or unknown. A job counts as remote if it offers a remote option
remoteScope	worldwide, regional, multi-country, single-country, not-remote or unknown
remoteScopeLabel	Human-readable summary, e.g. Worldwide, EMEA, Germany only, Hybrid: United Kingdom
isWorldwide	true if the job can be done from (almost) anywhere. eligibleCountries is then empty
eligibleCountries	ISO 3166-1 alpha-2 codes of the countries you may work from, with regions expanded
eligibleRegions	Regions named in the posting, e.g. EMEA, EU, LATAM
excludedCountries	Countries the posting explicitly excludes
eligibilityConfidence	high, medium, low, or null if the posting does not say (see below)
eligibilityEvidence	The texts the decision is based on: location, structured (ATS field), title or description
timezoneRequirement	Sentence stating a time-zone requirement, if any
employmentType	full-time, part-time, internship, temporary or unknown
engagementType	employee, contractor, employer-of-record or unknown
postedAt, postedAtSource	Publish date (Greenhouse: first publish date; Lever: created; Ashby: published)
salary	Structured salary range if the ATS publishes one (Lever, Ashby)
matchedCountries	Which of your requested countries this job is open to
descriptionText / descriptionHtml	Description as plain text and/or sanitised HTML

For onsite and hybrid jobs, eligibleCountries are the countries of the offices: you need to live there.

How eligibility is detected

The Actor reads three sources, in order of reliability:

1. The location field(s) of the posting, for example Remote - USA, Home based - EMEA, Remote, Canada; Remote, United States, SF, NYC, Remote. It recognises all countries, common regions (EU, EEA, Europe, EMEA, DACH, Nordics, CEE, LATAM, North America, Americas, APAC, ...), US states, Canadian provinces and about 400 tech-hub cities, with handling for ambiguous names ("London, Ontario", "Atlanta, Georgia" vs. "Tbilisi, Georgia", "Berlin, DE" vs. Delaware).
2. Structured ATS fields, such as Ashby's address country, Lever's country or a Greenhouse "remote countries" custom field.
3. Explicit sentences in the description, for example "This role is open to candidates based in the UK or Ireland" or "Candidates outside the US will not be considered".
4. The job title, only for a bare Remote location, and only if a segment names nothing but places: Salesforce Lead - USA, Engineer (United Kingdom). This gives medium confidence.

A job counts as remote only if the posting says so for this job. Company-wide statements such as "we are a distributed company" do not count, because they also appear on office jobs. Talent pools and general applications get low confidence, so they do not pass a default country filter.

Rules that keep results trustworthy:

• The description cannot widen a stated location. A job listed as Remote - USA stays US-only even if the company's boilerplate says "we work from anywhere in the world". If the location is just Remote, an explicit sentence about this role ("You must be based in Germany", "You can work from anywhere in the world") fills it in with medium confidence. Sentences about the company ("our team works remotely from Brazil to Japan") are ignored.
• "Anywhere in X" means X. Remote - Anywhere in the US is US-only, not worldwide.
• Inferences are marked, not hidden. If nothing states where the job may be done from, eligibilityConfidence is null. Weaker inferences, such as a remote job whose only hint is the company's office list, get low. Both are excluded from country-filtered results (and not charged) unless you lower minConfidence or enable Include jobs with unknown location rules.
• Contradictions are flagged, not hidden: if the location says Singapore but the text says "This role is based in San Francisco", confidence is low.

Confidence levels

Level	Meaning	Examples
high	Stated explicitly	Remote - Germany, Remote - EU, Remote - Anywhere in the US, an explicit remote-country list, a location confirmed by a "must be located in France" sentence
medium	Stated, but broad or from a single source	Broad regions such as EMEA, APAC or LATAM, a structured ATS country for a remote job, a remote option listed next to cities, a bare Remote plus one sentence in the description
low	Inferred or contradictory	A remote job whose only hint is the company's office list, Global (US-based preferred), location and description naming different countries

Broad regions are taken literally. EMEA covers Europe, the Middle East and Africa (116 countries), but many employers mean mainly the UK and EU. If you search from a country at the edge of such a region, check eligibleRegions and eligibilityEvidence, or set minConfidence to high to get only explicitly stated countries and precise regions (EU, EEA, DACH, Nordics, ...).

By default, country filters accept medium and better (minConfidence).

Regions and sanctions

Regions are business regions, not strict geography, and deliberately conservative: North America means US and Canada; Europe and EMEA do not include Russia or Belarus. Comprehensively embargoed countries (Cuba, Iran, North Korea, Syria) never match, not even for worldwide jobs.

Eligibility here means "you may be based there for this job". It does not cover visas or work permits.

Pricing

Pay per result: you pay only for jobs written to the dataset, which means only for jobs that pass all your filters. Fetching and filtering out non-matching jobs costs nothing. Use Maximum number of jobs or the run's spending limit to cap costs. The run stops cleanly when the limit is reached. If the platform restarts a run, jobs already stored are not stored or charged again.

Performance

All three sources are public JSON APIs, with no browser and no proxies. The built-in list of 240+ companies (about 20,000 jobs) took about 47 seconds with 4 GB of memory and about 3 minutes with 1 GB, because the platform assigns CPU in proportion to memory. A list of a few dozen companies is correspondingly faster.

Limitations

• Only public boards on Greenhouse, Lever and Ashby. Workday, SmartRecruiters, Personio and other ATS are not covered yet.
• Greenhouse publishes salary ranges only per job, so salary is empty for Greenhouse in this version.
• Greenhouse has no standard field for the employment type. It is read from the title and from explicit sentences about the role, and about two thirds of Greenhouse jobs state nothing usable: their employmentType and engagementType are unknown. Filtering on these two fields therefore drops those jobs.
• Greenhouse boards hosted on the EU instance are not supported yet.
• Detection is rule-based. Job postings are written by humans, so use eligibilityEvidence to double-check edge cases.
• excludeKeywords checks only the job title. keywords, departments and excludeKeywords match whole words: IT does not match Recruiting, and engineer does not match Engineering, so use department names as published.
• States or cities a posting excludes (Remote - US (excluding California), "we cannot hire in Alaska or Hawaii") are not modelled: the country stays eligible with medium confidence and the exclusion is listed in eligibilityEvidence.

Data and privacy

The Actor reads only public job postings published by companies for applicants. It does not log in, does not bypass any protection and does not collect personal data. Requests go only to the official public APIs (boards-api.greenhouse.io, api.lever.co, api.eu.lever.co, api.ashbyhq.com); company URLs you enter are never fetched directly. Job descriptions in descriptionHtml are sanitised (no scripts, styles, iframes, images, forms or event handlers; links are restricted to http/https/mailto), so they are safe to render on your site. All other fields (title, locations, descriptionText, ...) are plain text as published by the company and may contain characters such as <: escape them when you insert them into HTML.

FAQ

My company is not found. Check the slug in the board URL, e.g. jobs.ashbyhq.com/<slug> or job-boards.greenhouse.io/<slug>. Use a prefix (lever:<slug>) if a slug exists on several ATS. RUN_REPORT lists the status of every company.

Why is a job I expected missing? Most often the posting does not state where you may work from (eligibilityConfidence: null). Enable Include jobs with unknown location rules to see those too.

Can I get every job without filtering? Yes. Leave all filters empty and you get every job with its eligibility fields.