ATS Jobs + Tech-Stack & Hiring-Signal Intelligence

Build a normalized job-intelligence feed from public ATS job boards. Provide company targets or ATS board slugs, and this Actor fetches public postings, normalizes them into one schema, detects technical stack signals, scores AI/infra relevance, and can emit only job changes between runs.

It is built for investors, GTM teams, recruiters, founders, talent-intelligence workflows, and agentic research tasks that need more than raw ATS records.

Supported ATS platforms

• Greenhouse
• Lever
• Ashby
• SmartRecruiters, when the company's public Posting API feed is enabled
• Workable, through the public careers widget JSON feed when available

Workday and Personio are intentionally not included in v0.1.

What makes it different

• One unified output schema across multiple ATS sources.
• Deterministic tech-stack enrichment: no LLM runtime calls and no unpredictable model cost.
• AI, ML, infrastructure, platform, data, DevOps, security, product, and non-engineering role-family detection.
• Relevance scores designed to separate strong AI/infra matches from generic engineering roles.
• Delta mode for new, changed, removed, and optionally unchanged job events.
• Curated company registry for common public boards, while still accepting explicit ATS slugs.

Input example

{
  "targets": [
    {
      "company": "Databricks",
      "ats": "greenhouse",
      "slug": "databricks"
    },
    {
      "company": "Shield AI",
      "ats": "lever",
      "slug": "shieldai"
    },
    {
      "company": "OpenAI",
      "ats": "ashby",
      "slug": "OpenAI"
    }
  ],
  "keywords": ["AI", "ML", "machine learning", "LLM", "Kubernetes", "GPU", "inference"],
  "maxJobsPerCompany": 100,
  "maxItems": 500,
  "minRelevanceScore": 25,
  "includeGenericEngineering": true,
  "includeDataPlatform": true,
  "includeNonEngineering": false,
  "locations": [],
  "postedOrUpdatedWithinDays": 365,
  "mode": "full",
  "deltaKey": "default"
}

If mode is present, it is canonical. enableDeltaMode=true is only a compatibility shortcut when mode is omitted.

Output example

{
  "source": "greenhouse",
  "ats": "greenhouse",
  "company": "Databricks",
  "companySlug": "databricks",
  "jobId": "7890123",
  "internalJobId": "REQ-4567",
  "title": "Senior ML Infrastructure Engineer, Inference Platform",
  "department": "Engineering",
  "team": "AI Platform",
  "location": "San Francisco, CA",
  "office": "San Francisco",
  "remoteType": "hybrid",
  "workplaceType": "hybrid",
  "employmentType": "Full-time",
  "seniority": "senior",
  "roleFamilies": ["ai_infrastructure", "mlops", "gpu_platform", "inference_serving", "kubernetes_platform"],
  "detectedSkills": ["Python", "Kubernetes", "GPU", "CUDA", "MLOps", "vLLM", "Triton"],
  "detectedClouds": [],
  "detectedFrameworks": [],
  "detectedDataTools": [],
  "detectedInfraTools": ["Kubernetes", "GPU", "CUDA", "vLLM", "Triton"],
  "salaryMin": 160000,
  "salaryMax": 220000,
  "salaryCurrency": "USD",
  "salaryRaw": "$160,000 - $220,000",
  "postedAt": "2026-06-24T17:12:00Z",
  "updatedAt": "2026-07-02T09:30:00Z",
  "absoluteUrl": "https://job-boards.greenhouse.io/databricks/jobs/7890123",
  "contentText": "Build GPU inference platforms on Kubernetes with CUDA, vLLM, Triton, Python, and MLOps workflows.",
  "relevanceScore": 94,
  "aiInfraScore": 96,
  "confidence": "high",
  "whyItMatches": "Matches ML infrastructure, GPU, CUDA, vLLM, Triton with ai infrastructure, mlops, gpu platform role signals.",
  "hiringSignalTags": ["strong_fit", "gpu_hiring", "inference_hiring", "kubernetes_platform_hiring", "ai_infra_stack"],
  "isNew": false,
  "isChanged": false,
  "isRemoved": false,
  "changeType": "full",
  "previousHash": "",
  "currentHash": "a8d9f8b2e18f88d4a67f9f74d8c3b4e63873a8f0bff26795c8f8d3f0a8b9c123",
  "firstSeenAt": "2026-07-07T13:10:04Z",
  "lastSeenAt": "2026-07-07T13:10:04Z",
  "runTimestamp": "2026-07-07T13:10:04Z"
}

team may be empty for ATS sources that do not expose team data.

Full mode vs delta mode

mode=full emits all matching enriched records for the current run.

mode=delta stores a prior snapshot in a named Actor key-value store, scoped by this Actor, runtime user/task context when Apify provides it, and deltaKey. It compares the current run with that snapshot and emits:

• new_job
• changed_job
• removed_job
• unchanged, only when emitUnchangedInDeltaMode=true

Delta baseline behavior: On the first delta run for a given deltaKey, all matching current jobs are emitted as new_job and saved as the baseline. Later runs with the same scope and deltaKey emit only new_job, changed_job, removed_job, and optionally unchanged when emitUnchangedInDeltaMode=true.

For scheduled monitors or multiple saved tasks, use a unique deltaKey per monitor/task. Apify task/user context is included when available, but unique keys avoid accidental collisions in manual runs or custom orchestration.

Source policy and disclaimer

This Actor uses public, unauthenticated job-board JSON endpoints only. It does not log in, use browser automation, use proxies, call paid APIs, submit applications, enrich personal data, or scrape candidate/recruiter private data.

This Actor is not affiliated with Greenhouse, Lever, Ashby, SmartRecruiters, Workable, or any hiring company. Availability and fields depend on each company's public ATS configuration.

Limitations

• Public ATS fields vary by company and platform.
• Salary is parsed only when visibly present in public data.
• SmartRecruiters and Workable feeds may not be enabled for every company.
• Delta mode compares jobs after input filtering, so removed events refer to jobs that previously matched the same task configuration.

Pricing suggestion

Suggested starting price later: $2 per 1,000 full records. A future delta/change-event tier could be $4-5 per 1,000 delta events. This repository change does not configure monetization or publish the Actor.