O*NET Skill & Software Records — Occupation Associations

Official source. No API key. Pay per record.

Who this is for and the decision it supports

HR-tech, workforce-planning, skills-taxonomy and job-matching teams who need the official O*NET links between occupations and the skills and software they use.

One record answers: which skill or software tool is associated with an occupation, and how strongly — a building block for skills graphs, job-to-skill mapping, and reskilling models.

What one record is

Each record is one occupation-to-element association: O*NET-SOC code and title, the skill or software element, its class, scale/value, and hot-technology / in-demand flags.

Your first run

Paste this input and run it:

{
  "onetSocCodePrefix": "15-1252",
  "associationType": "software",
  "maxRecords": 10
}

Filter by an occupation code prefix and association type and get clean association rows immediately. Here is a real record from that exact input:

{
  "record_id": "15-1252.00:software:2.E.4.d",
  "association_type": "software",
  "onet_soc_code": "15-1252.00",
  "occupation_title": "Software Developers",
  "skill_class": null,
  "element_id": "2.E.4.d",
  "element_name": "Word processing software",
  "scale_id": null,
  "data_value": null,
  "software_example": "3M Post-it App",
  "hot_technology": false,
  "in_demand": false
}

What the sample costs: $0.05 per record. The sample above delivers 10 records, so it costs at most $0.50.

Pulling in bulk

Drop the filters and raise Maximum records to pull the whole occupation-skill-software matrix. A buyer pulls thousands of records to load the full occupation-to-skill/software matrix into a taxonomy or matching engine.

To expand from the sample: raise Maximum records, and widen any filters shown in the input. You are billed only for records actually delivered, so a sample stays cheap and a full pull is predictable.

When to run it again

O*NET republishes its database on a scheduled release; re-run after a new release to refresh the matrix.

Update schedule: ONET publishes database updates on a regular release schedule (multiple updates per year in recent cycles). Verified live: association rows return with current ONET element identifiers and hot-technology flags.

How this is different

Occupation-to-skill/software associations from O*NET specifically — distinct from job-posting feeds; this is the authoritative taxonomy layer, not live vacancies.

Limitations (source coverage vs product coverage)

• This is the O*NET taxonomy (occupation to skill/software), not live job postings or salaries.
• Coverage is US O*NET-SOC occupations.
• Software examples are O*NET's named examples, not an exhaustive tool inventory. Each software row is individually addressable: the record identity now includes the specific tool (Workplace Example), so different tools under one technology category are distinct records and exact duplicates are removed before delivery and billing.

FAQ

Is this job postings?

No. It is the O*NET occupation-to-skill/software taxonomy. For live vacancies use a job-postings feed.

What is a Hot Technology?

O*NET's flag for software/technology in high employer demand; filter with hotTechnologyOnly.

What does an empty run mean?

No association matched your filters — widen the occupation prefix or association type.

Related actors

• world-bank-project-records — development-sector programs
• eu-business-demography-records — firm demography for labour-market context
• company-data-enrichment-api — resolve employers to legal entities

Source, licence and how it is built

Source and licence are carried on every record (source, licence, attribution, observed_at). The actor reads only the official public endpoint, honours the source's robots and licence terms, and emits values unmodified with full provenance. Records are keyed on a stable id so re-runs upsert cleanly into a table you keep. Row delivery and billing are exactly equal: you are charged once per record delivered (push-then-charge), so an empty run costs nothing.

Part of NexGen Signal — official-source data products.