Wellfound Jobs Scraper - Startup Jobs

Collect public startup job listings from Wellfound, formerly AngelList Talent, and export them as structured recruitment and market-intelligence data. Search one or more locations, match role or skill keywords, filter remote jobs and compensation, and receive full public descriptions plus startup metadata in every result.

The Actor returns job titles, descriptions, salary and equity ranges, locations, remote policy, experience requirements, startup name, size, profile links, logos, hiring badges and publication times. Results can be downloaded as JSON, CSV, Excel or XML, scheduled in Apify, or consumed through the dataset API.

What can you use this Actor for?

• Monitor startup and technology hiring by city, country or region.
• Find roles that disclose salary or equity before applying.
• Build startup hiring-signal lists for sales, recruiting or investment research.
• Compare remote-work availability across startup markets.
• Track companies growing in engineering, data, product, sales or operations.
• Create a job feed, alerting workflow, spreadsheet or labor-market dashboard.
• Analyze compensation bands, experience requirements and startup size.
• Discover recently published roles without manually checking location pages.

Why scrape Wellfound jobs?

Wellfound focuses on startups and technology companies. Its public job records often contain data that general job boards omit: salary bands, equity ranges, accepted remote locations, startup size, role category and company positioning. That makes the source useful for candidates, recruiters, analysts, founders and B2B teams.

The Actor reads the structured Apollo state that Wellfound publishes on its public location pages. A single page can contain dozens of job records with complete descriptions and the associated startup entities. This direct HTTP design avoids a full browser and does not require a Wellfound account, paid proxy or third-party API key.

Input

The default run searches Berlin for software engineer and returns up to 10 jobs. Use country-scale pages such as United States for broader research, or city pages such as Berlin, London and New York for a focused dataset.

Field	Purpose
locations	Wellfound location pages to collect, expressed as readable city, country or region names.
keywords	Terms matched against title, description, role, startup name and company summary.
remoteOnly	Keep only jobs marked remote or remote-capable by Wellfound.
jobTypes	Optional full-time, part-time, contract or internship filters.
minimumSalary	Minimum lower bound of the published salary range.
includeJobsWithoutSalary	Decide whether jobs without disclosed compensation remain in output.
maxResultsPerLocation	Maximum unique matching jobs saved for each location.
maxPagesPerLocation	Safety cap for Wellfound pagination.

Example input

{
  "locations": ["Berlin", "United States"],
  "keywords": ["python", "machine learning"],
  "remoteOnly": true,
  "jobTypes": ["full-time"],
  "minimumSalary": 100000,
  "includeJobsWithoutSalary": false,
  "maxResultsPerLocation": 25,
  "maxPagesPerLocation": 10
}

Keywords use an OR match: a row is included when any requested term appears in its title, full public description, primary role, startup name or company summary. Jobs that appear in more than one location are deduplicated by Wellfound job ID.

Output

Each dataset item is one public startup job. Common fields include:

• job_id, url, title, description
• job_type, primary_role
• locations, accepted_remote_locations
• is_remote, remote_kind, work_from_home_flexible
• years_experience_minimum, years_experience_maximum
• compensation, salary_minimum, salary_maximum, salary_currency
• equity_minimum_percent, equity_maximum_percent
• published_at, is_auto_posted, ats_source
• company_id, company_name, company_url, company_logo
• company_size, company_high_concept, company_badges
• lead_score, source_location, search_keywords
• search_rank, source_url, scraped_at

Example result

{
  "job_id": "4420215",
  "url": "https://wellfound.com/jobs/4420215-middle-data-engineer",
  "title": "Middle Data Engineer",
  "job_type": "full-time",
  "primary_role": "Data Engineer",
  "locations": ["Berlin", "Rome", "Madrid"],
  "accepted_remote_locations": ["Europe"],
  "is_remote": true,
  "compensation": "$35k – $48k • No equity",
  "salary_minimum": 35000,
  "salary_maximum": 48000,
  "salary_currency": "USD",
  "years_experience_minimum": 3,
  "company_name": "UVIK Software",
  "company_size": "SIZE_11_50"
}

Compensation and equity

The Actor preserves Wellfound's original compensation text and also parses common dollar, euro and pound ranges into numeric fields. Values with k or m suffixes are normalized to complete numbers. Equity percentages are returned separately when the employer publishes them.

The salary_currency field is essential when combining several markets. Do not compare numeric salary values across currencies without conversion. Missing compensation stays empty, and the Actor never invents or estimates a salary.

Startup hiring intelligence

Every job is joined to the startup entity published on the same page. The output can include the startup name, public profile, logo, team-size band, one-line positioning and badges such as actively hiring, recently funded, fast growing or top responder.

lead_score is a transparent completeness score, not a prediction. It rewards rows with a named startup, full description, compensation, equity, logo, company positioning, locations and role metadata. Use it to prioritize records for manual review, not to automate consequential decisions.

Reliable extraction design

The Actor uses browser-compatible HTTP fingerprints to fetch public location pages, then reads their Next.js Apollo state. It does not depend on fragile visual selectors. Pagination is bounded, overlapping jobs are deduplicated by source ID, and temporary failures are retried with short backoff.

Runs fail visibly when no requested job matches are found. This avoids presenting an empty dataset as a healthy extraction. Defaults are intentionally small enough for a quick first run and Apify's automated quality tests, while production limits remain configurable.

Scheduling and integrations

Run the Actor daily to monitor a stable list of locations and keywords. Use job_id as the external primary key and alert only when a new value appears. A recruiting team can route new matches to Slack or email; an analyst can append snapshots to a warehouse; a founder can monitor competing startups' hiring activity.

Apify datasets can be exported directly or connected to Make, Zapier, n8n, webhooks and cloud storage. Use CSV or Excel for ad hoc research and JSON for recurring applications.

Pricing

This Actor uses transparent pay-per-event pricing. A tiny start event covers initialization, and each unique job pushed to the default dataset is billed as one result. Records rejected by filters or removed as duplicates are not billed as dataset results.

The result price is positioned below the current high-usage Wellfound competitors while retaining full descriptions, compensation, equity and startup context.

Responsible use

The Actor collects public job and company information. Use it for legitimate job search, recruitment, research, monitoring and integration workflows. Follow applicable privacy, employment, marketing and database rules. Verify current details on the canonical Wellfound page because jobs can change or expire after collection.

FAQ

Does this require a Wellfound login?

No. The Actor reads public location-page data without a user account.

Can I search by role?

Yes. Add role or skill terms to keywords. They are matched against the title, primary role and full description as well as startup text.

Can I collect remote startup jobs?

Enable remoteOnly. Inspect remote_kind and accepted_remote_locations for the source's more detailed remote restrictions.

Why are some salary or equity values empty?

Employers choose which compensation fields to publish. Keep includeJobsWithoutSalary disabled when your workflow requires a disclosed salary.

How do I search New York or the United States?

Enter readable values such as New York or United States. The Actor converts them to Wellfound location slugs.

Why did a very narrow search return no rows?

The requested keywords and filters may not match the pages scanned. Increase maxPagesPerLocation, broaden the keywords, allow jobs without salary, or remove the remote-only filter.

How do I deduplicate scheduled runs?

Use job_id as the primary key and url as a fallback. The Actor already removes duplicates within each run.

How can I report a source change?

Open an issue from the Actor page and include a small input plus the failed run URL. Do not include login details or private candidate information.

You might also like

• Welcome to the Jungle Jobs Scraper for European startup and employer-profile jobs.
• Y Combinator Jobs Scraper for YC startup vacancies.
• Hiring Cafe Scraper for broad technology job discovery.
• Remote Jobs Aggregator for multi-source remote roles.

If this Actor saves you time, please leave a review on its Apify Store page. Reviews help prioritize reliability fixes and requested filters.

Keywords: Wellfound scraper, Wellfound jobs scraper, AngelList jobs API, startup jobs scraper, startup hiring data, remote startup jobs, tech jobs API, salary equity data, startup lead generation, recruitment intelligence, Wellfound API, Apify jobs scraper

Production workflow

This Actor is designed for job seekers, startup recruiters, labor-market analysts and sales teams monitoring hiring signals. Schedule focused location and role searches, deduplicate by job_id, and compare each dataset with the previous run to identify new openings, compensation changes and companies accelerating hiring. Small, repeatable runs are easier to audit than occasional unlimited exports: they expose source changes quickly, reduce duplicate processing and keep automation costs predictable. Use the input defaults as a smoke test, then widen the query only after confirming that the returned fields match the downstream workflow.

Treat each dataset row as source evidence captured at a point in time, not as a permanent fact. Preserve the canonical identifier and source URL, add the run timestamp in the receiving system, and deduplicate before writing to a CRM, spreadsheet, warehouse or alert channel. For recurring monitoring, compare the newest dataset with the previous successful run and send only additions or material changes onward. That pattern avoids repeated notifications and makes the automation more useful over time.

Data quality and interpretation

The primary output includes job_id, url, title, description, job_type, primary_role, locations, accepted_remote_locations, is_remote, remote_kind, work_from_home_flexible, years_experience_minimum. Availability depends on what the source publishes for a particular result. Empty values are retained as null or empty fields rather than invented. Search rank is source-specific and can change with language, geography, query wording and time. Validate critical decisions against the linked source page, especially when the data is used for outreach, compliance, pricing, routing or reporting.

Use bounded inputs and respect the source's published interface. The Actor does not bypass logins or collect private account data. Avoid placing passwords, session cookies, personal access tokens or confidential records in Actor input. If a public source changes its response shape, attach a reproducible public input and run ID to an Actor issue so the parser can be updated without exposing private data.

Automation and cost control

The default dataset can be exported as JSON, CSV, Excel, XML or RSS and can be consumed through the Apify API. In Make, Zapier or n8n, start the Actor, wait for the run to finish, then read default-dataset items. In code, store the Actor run ID and dataset ID alongside the imported rows so every record remains traceable. Apply filters after collection only when the source cannot express them directly; source-side filtering usually reduces cost and noise.

This Actor uses pay-per-event billing. A result event is charged only when a structured row is saved to the default dataset. Keep test runs small, estimate expected row volume before scheduling broad searches, and use platform spending limits for unattended workflows. Failed requests and empty searches should not create result events.

Support and feedback

If the Actor saves time in a real workflow, please leave a short Apify Store review describing the use case and the fields that were most useful. Reviews help other users evaluate the product and help prioritize maintenance. For support, open an Actor issue with the run ID, public input, expected behavior and one affected field. Do not include credentials or confidential information.