Job Posting AI Analyzer

Scrapes job posting pages and runs each through Claude Haiku to extract structured sales intelligence: seniority level, required skills, culture signals, company size signal, and a hiring signal score (1–10) indicating how strong a B2B sales prospect this company is.

What it does

1. Takes a list of job posting URLs (Greenhouse, Lever, Ashby, Workday, or any public page)
2. Scrapes each page — title, company, location, salary if present, description
3. Sends the content to Claude Haiku (claude-haiku-4-5) for AI analysis
4. Pushes one structured row per job to the Apify dataset

Use case: B2B sales prospecting

B2B sales teams use job postings as buying signals. A company hiring a Head of Revenue Operations is almost certainly evaluating new tooling. A startup posting five engineering roles at once is in a growth phase — revenue decisions are live. This actor turns a list of job URLs into a scored, enriched dataset you can pipe directly into your CRM or outreach sequence.

Feed it your target accounts' active job listings. Filter by hiringSignalScore >= 7. Use suggestedOutreachAngle to personalize the first touch.

Input

{
  "jobUrls": [
    "https://job-boards.greenhouse.io/anthropic/jobs/5101378008",
    "https://job-boards.greenhouse.io/anthropic/jobs/5208218008",
    "https://jobs.lever.co/acme/some-uuid"
  ],
  "anthropicApiKey": "sk-ant-...",
  "maxJobs": 10
}

Field	Type	Required	Description
jobUrls	string[]	Yes	Job posting URLs to process
anthropicApiKey	string	Yes	Anthropic API key (kept secret)
maxJobs	integer	No	Cap on URLs to process, default 10, max 100
proxyConfig	object	No	Apify proxy configuration

Output fields

Field	Type	Description
url	string	Original job posting URL
jobTitle	string	Extracted job title
company	string	Company name
location	string	Location or remote info
salaryRange	string|null	Salary if present on page
postedDate	string|null	Post date if detectable
descriptionSnippet	string	First 500 chars of job description
scraped	boolean	Whether the page was successfully scraped
scrapeError	string|null	Error message if scrape failed
enrichment.seniorityLevel	string	Junior / Mid / Senior / Lead / Executive
enrichment.requiredSkills	string[]	Must-have skills extracted from JD
enrichment.niceToHaveSkills	string[]	Preferred but not required skills
enrichment.cultureSignals	string[]	Culture/values signals from the JD
enrichment.hiringSignalScore	integer	1–10, how strong a B2B buying signal this role represents
enrichment.hiringSignalReason	string	One sentence explaining the score
enrichment.companySizeSignal	string	Startup / SMB / Mid-market / Enterprise (inferred from JD)
enrichment.suggestedOutreachAngle	string	How a B2B vendor should open with this company

Notes

• Greenhouse and Lever job URLs are resolved via their public APIs for clean data extraction. Other URLs are scraped with CheerioCrawler.
• The actor respects maxJobs — if you pass 50 URLs with maxJobs: 10, only the first 10 are processed.
• Anthropic API costs are separate from Apify platform costs. Claude Haiku is low-cost; expect under $0.01 per job posting at typical JD lengths.