B2b Ats Open Jobs Tech Stack Scraper

Turn company career pages into open jobs, salary ranges, and technology buying signals in one structured dataset.

Monitor Greenhouse, Lever, Ashby, Workable, and SmartRecruiters using public HTTP APIs. Start with company domains or known ATS boards, filter for departments or remote roles, and export results directly into your prospecting and automation workflows.

The Actor scans full job descriptions for more than 120 enterprise technology signatures, including Rust, Go, TypeScript, Python, Kubernetes, Terraform, AWS, GCP, Azure, Snowflake, Databricks, Kafka, Postgres, React, Next.js, Salesforce, HubSpot, NetSuite, Workday, and SOC2. A mention is a hiring signal, not proof of a product purchase or deployment.

Best for recruiting research, hiring signals and account segmentation. Each successful row is an open job; technology mentions describe hiring requirements, not confirmed software purchases. Use a direct ATS board URL for the most predictable first run.

Ready-to-run example

The Input form includes public example values. A fresh apify run uses the input schema defaults. To explicitly select the small sample, including when input was saved by an earlier local run, use:

apify run --input-file .actor/input.example.json

You can also paste .actor/input.example.json into the JSON input editor in Apify Console. Replace the example sources with your own targets and increase the limits when you are ready.

How to use

1. Add company domains, direct ATS board URLs, or provider identifiers to Company domains or ATS boards.
2. Optionally select a department substring, remote roles only, and the maximum number of matching jobs per company.
3. Start the run. Review the Open jobs table, then export JSON, CSV, Excel, or retrieve items through the dataset API.

Example target	Behavior
https://boards.greenhouse.io/stripe	Read the Stripe Greenhouse board
https://jobs.ashbyhq.com/linear	Read the Linear Ashby board
lever:palantir	Read a known Lever board
workable:huggingface	Read a known Workable board
smartrecruiters:BoschGroup	Read a known SmartRecruiters board
https://linear.app	Discover the company's ATS from its website

Corporate discovery checks /careers, /jobs, and the supplied page. It recognizes links, embedded boards, widget attributes, and embedded ATS URLs. If no link is found, it probes the supported providers using the domain's company name. Explicit board URLs are best for companies whose ATS slugs differ from their domains.

Use examples/input.json for a two-row quick run:

{
  "targets": [
    "https://boards.greenhouse.io/stripe",
    "https://jobs.ashbyhq.com/linear"
  ],
  "maxJobsPerCompany": 1,
  "maxConcurrency": 20,
  "timeoutSecs": 15
}

The Console prefill includes two boards with three jobs per company, for up to six rows. The schema default is also three jobs per company. Increase maxJobsPerCompany for larger runs. Public API latency and availability determine run time; a particular company's availability or a sub-five-second run cannot be guaranteed.

What data can you extract?

Every row has the same flat structure. Missing public fields are empty strings; detectedTechStack is an array and isRemote is a boolean.

Field	Meaning
companySlug	ATS board slug, or public target identifier on discovery failure
atsPlatform	greenhouse, lever, ashby, workable, or smartrecruiters
status	SUCCESS or ERROR
jobId	Provider identifier, with a stable derived fallback when omitted
jobTitle	Published position title
department	Public department or team labels
location	Published location
isRemote	Whether the job is explicitly remote
workplaceType	Remote, Hybrid, OnSite, or empty when unknown
employmentType	Provider's employment label
publishedAt	Public publication date in UTC ISO 8601, when supplied
salaryRangeText	Structured compensation or a salary range from the description
detectedTechStack	Distinct technology mentions from the full description
descriptionSnippet	Plain-text description prefix, up to 1500 characters
applyUrl	Public application or job URL
scrapedAt	UTC extraction timestamp
errorMessage	Safe diagnostic text and code for errors

The Actor preserves disclosed currency and compensation periods and does not estimate missing salaries. Lever exposes a creation timestamp, which is used for publishedAt. Explicit hybrid and on-site labels take precedence over remote location wording. Descriptions are not used to infer remote eligibility.

Input settings

Setting	Default	Accepted values
targets	Required	1 to 10000 public domains, board URLs, or provider identifiers
departmentFilter	Empty	Case-insensitive literal department substring, up to 200 characters
onlyRemote	false	Include explicitly remote jobs; exclude hybrid and unknown workplaces
maxJobsPerCompany	3	Integer from 1 to 1000; Console prefill is 3
maxConcurrency	20	Integer from 1 to 50, shared by every HTTP request
timeoutSecs	15	Integer from 3 to 60 per request attempt
proxyConfiguration	Direct first	Optional Apify proxy editor configuration

Filters run before the matching-job limit, and pagination continues through non-matching jobs. Duplicate targets and company boards are consolidated within a run. Error rows do not use the successful-job allowance, but both row types count toward the free-plan cap.

Omit proxy settings for direct requests with datacenter fallback after HTTP 403, 429, or 503. The Actor tries SHADER, then the account's default datacenter proxy. A configured final tier is used only with explicit useApifyProxy: true. Custom proxyUrls force those proxies; useApifyProxy: false forces direct connections. Automatic Apify tiers require platform proxy configuration. Local runs use direct requests unless custom proxies are supplied.

Output example

This illustrative row shows the output shape:

{
  "companySlug": "example",
  "atsPlatform": "ashby",
  "status": "SUCCESS",
  "jobId": "engineering-platform-2026",
  "jobTitle": "Senior Platform Engineer",
  "department": "Engineering",
  "location": "Remote - United States",
  "isRemote": true,
  "workplaceType": "Remote",
  "employmentType": "FullTime",
  "publishedAt": "2026-10-01T12:00:00.000Z",
  "salaryRangeText": "$150,000 - $200,000 per year",
  "detectedTechStack": ["AWS", "Kubernetes", "Python", "Terraform"],
  "descriptionSnippet": "Build our platform with Python, Kubernetes, Terraform and AWS. Salary: $150,000 - $200,000 per year.",
  "applyUrl": "https://jobs.ashbyhq.com/example/engineering-platform-2026/apply",
  "scrapedAt": "2026-10-19T12:00:00.000Z",
  "errorMessage": ""
}

Failed targets produce ERROR rows with a public identifier and a diagnostic such as HTTP_404, ETIMEDOUT, or ATS_NOT_FOUND. Valid boards with no matching jobs produce zero job rows. Totals and the stopping reason are saved as RUN-SUMMARY in the default key-value store.

Use the results in Clay, n8n, Make, or your API

In Clay, import the dataset or CSV, group by companySlug, and use technology mentions and department hiring as enrichment signals. Select status = SUCCESS before building prospect segments.

In n8n or Make, use the Apify integration to run a saved task, wait for completion, and retrieve dataset items. Compare companySlug, atsPlatform, and jobId with your previous run to identify new or removed postings. Schedule recurring monitoring in Apify Console.

For API workflows, obtain the completed run's defaultDatasetId and call the dataset items endpoint, selecting JSON or CSV with format. Authenticate through the integration's connection or authorization header. Keep credentials out of target URLs and exported data.

Pricing and limits

The custom pay-per-event name is job-scraped. Only saved SUCCESS rows are submitted with this event. ERROR rows cost $0 in custom event fees. Platform compute, storage, transfer, and any proxy charges can still apply.

The price is configured in Apify Store settings. For example, at $0.001 per successful job, 1000 successful jobs cost $1.00 in event fees; an additional 100 error rows add $0.00 in custom event fees. This calculation is illustrative; check the Store's displayed price before running.

On the platform, non-paying users can save 25 total rows per run, including errors. Paid users and local runs do not have that cap. The writer reserves the remaining custom-event budget across buffered results. Reaching either limit stops new work and exits successfully. Results flush at 100 rows, every five seconds, and during shutdown.

The container defaults to 256 MB and a one-hour timeout. Decompressed responses are capped at 2 MiB for JSON, 4 MiB for HTML, and 15 MiB for XML. Small Greenhouse runs use lean listings and selected details; oversized full feeds fall back to that approach. Lever uses adaptive pagination, Workable uses cursor tokens, and SmartRecruiters uses offsets and detail requests. A board is limited to 50000 scanned postings.

FAQ and troubleshooting

Why is salary or publication date empty? The provider may not publish it. Only disclosed salary values or recognizable ranges are extracted. Updated timestamps are not substituted for publication dates.

Why did a domain return ATS_NOT_FOUND? The site may require JavaScript, use another ATS, or use an unrelated slug. Supply a direct supported board URL. Discovery does not run a browser.

Can this extract private jobs? It reads public boards and descriptions. It does not log in, submit applications, or access recruiting records.

Why are fewer jobs returned? Filters, the free-plan cap, the event budget, or current openings can reduce the count. Failed details appear as uncharged errors. Empty boards do not produce synthetic successful records.

What happens on provider errors? Timeouts, TLS failures, blocked responses, and malformed payloads become error records so other companies can finish. Invalid input fails validation. An unavailable platform or output-storage service stops the run safely with a status message; output delivery cannot be guaranteed during a storage outage.

How are requests protected? Every destination and redirect is checked against public DNS addresses. Direct requests retain the original hostname while pinning validated IPs. Proxy tunnels connect to validated public IPs with the origin's TLS server name. Certificate verification remains enabled. Unsafe links, excessive nesting, prototype-pollution keys, and spreadsheet formulas are handled before writes.

Which APIs are used? The adapters use the Greenhouse Job Board API, Lever Postings API, Ashby's hosted GraphQL API with its documented public posting API for enrichment and fallback, Workable's public board endpoints, and the SmartRecruiters Posting API. Workable and SmartRecruiters detail calls supply descriptions absent from list responses.

Local development

Use Node.js 22 or newer:

npm ci --ignore-scripts
npm run build
npm test
npm run smoke
npm run smoke -- --all

Unit tests make no remote requests. Smoke tests make live read-only requests, requesting one job per board. --all adds Lever, Workable, and SmartRecruiters checks. Test and smoke subprocesses enable UV_THREADPOOL_SIZE=64, production mode, and strict TLS verification.

To run the complete Actor locally, save the example as storage/key_value_stores/default/INPUT.json, then run npm start. With Apify CLI, use apify run --input-file examples/input.json. See DEPLOYMENT.md for container and Store setup.

Repository initialized by Arezgit.