O*NET Skill & Software Records — Occupation Associations

Keyless. Pay per record. Every occupation's skills and software, as one clean association table.

This actor turns the O*NET database (the U.S. Department of Labor's occupational data system) into clean, per-association records — the map of which skills and which software each occupation uses. Each run reads ONET's keyless bulk database and delivers one row per occupation-to-element link: the occupation, the skill or software element, and the ONET rating or flags that describe it. Occupation-to-skill / occupation-to-software grain, structured for engineering use, under the Creative Commons Attribution 4.0 licence, with no personal data of any kind.

What this is, in one paragraph

ONET describes roughly a thousand occupations along hundreds of dimensions. Two of those dimensions matter most for skills intelligence: the skills an occupation requires (with ONET's importance and level ratings) and the software / technology an occupation uses (with ONET's hot-technology and in-demand flags). This actor flattens both into a single association table — one record per occupation-element pair — so you can answer "which occupations use Python," "what skills matter most for this role," or "how does this job's technology profile compare to that one" with a plain join instead of parsing ONET's file bundle by hand. Values are exactly as O*NET publishes them. You are billed only for records actually delivered.

Who buys this and for what job

• Workforce-analytics and talent-intelligence products that need occupation→skill and occupation→software mappings as clean rows to power role taxonomies, skill graphs, and gap analysis.
• HR-tech, job-matching, and resume-parsing tools aligning free-text roles to a standard skill and technology vocabulary.
• Labor-market researchers and economists studying skill demand, technology adoption, and how tooling differs across occupations.
• Data engineers who want the O*NET skills and technology-skills tables joined and normalized without downloading and reshaping the raw database bundle.

Creative Commons Attribution 4.0 — free commercial reuse

One licence rider travels on every record. The ONET database is published by USDOL/ETA under the Creative Commons Attribution 4.0 International (CC BY 4.0) licence — free to copy, publish, distribute, and adapt, including commercially, with the required attribution: "This product incorporates information from the ONET database, used under CC BY 4.0. O*NET is a trademark of USDOL/ETA." That attribution string is carried on every record and in this README so downstream users always have it. This is occupational reference data — a description of work, not a directory of workers.

Occupation grain — no person data

This actor works at occupation grain: the unit of every record is an occupation (an ONET-SOC code), never a person. ONET contains no natural-person data — it describes jobs, not job-holders — and this actor emits none. Occupation titles, skill and software element names, rating values, and O*NET flags are the entire payload; a person-key guard runs on every record as a pure backstop and the acceptance suite feeds a deliberately person-shaped row and asserts none of it can ever appear in output.

Pricing

Event	Free plan	Bronze	Silver	Gold / Platinum / Diamond
association_record	$0.05	$0.045	$0.04	$0.0335

Delivered before charged. Blocked runs and empty results cost nothing.

Input

Field	Type	Default	Description
associationType	string	all	all, skill, or software.
onetSocCodePrefix	string	—	Optional. Keep only occupations whose O*NET-SOC code starts with these digits (e.g. 15 for computer/math, 29-1141.00 for one occupation).
elementContains	string	—	Optional. Keep only associations whose element name contains this text (e.g. Python, welding, negotiation).
hotTechnologyOnly	boolean	false	Software only: keep only technologies O*NET flags as a Hot Technology.
maxRecords	integer	500	Ceiling on records delivered and billed.

Output

One JSON object per occupation-element association. Real examples (values as published by O*NET):

{
  "record_id": "15-1252.00:software:2.E.5.b:",
  "association_type": "software",
  "onet_soc_code": "15-1252.00",
  "occupation_title": "Software Developers",
  "skill_class": null,
  "element_id": "2.E.5.b",
  "element_name": "Web platform development software",
  "scale_id": null,
  "data_value": null,
  "software_example": "Python",
  "hot_technology": true,
  "in_demand": false,
  "source": "O*NET database (db_30_3), USDOL/ETA",
  "licence": "O*NET database ... CC BY 4.0 ...",
  "attribution": "This product incorporates information from the O*NET database, used under CC BY 4.0 ...",
  "disclaimer": "Occupation-to-skill and occupation-to-software association records ... no natural-person data ...",
  "observed_at": "2026-08-21T00:00:00Z"
}

A skill association instead carries scale_id (IM importance or LV level), a numeric data_value, and a skill_class of essential or transferable, with the software fields null. An unbilled RUN_RECEIPT records the robots check, the filters, rows seen and delivered, and whether charged equals delivered.

Field reference

Each record is one occupation-element association. Identity: record_id (SOC code + type + element id + scale), association_type (skill / software), onet_soc_code, and occupation_title. Element: element_id and element_name. Skill-only: skill_class (essential / transferable), scale_id (IM / LV), and data_value (the O*NET rating). Software-only: software_example (the technology example), hot_technology, and in_demand. Provenance: source, source_url, licence, attribution, disclaimer, and observed_at.

How a run works

A run first re-reads the source host's robots posture from your runtime — the ONET Resource Center — and confirms the database download path is allowed, stopping at the door if a Disallow ever covers it. It then fetches the ONET database bundle once, reads the skills and technology-skills tables, joins each row to its occupation title, and flattens it to one association record — reading only occupation, element, rating, and flag fields, never anything person-shaped — delivered before it is charged, so a blocked or empty run costs you nothing. Your associationType, onetSocCodePrefix, elementContains, and hotTechnologyOnly filters are applied before delivery, and the actor stops at your maxRecords ceiling and records charge_equals_delivered in the receipt.

Coverage and scope

The database holds on the order of 94,000+ occupation-element associations across roughly a thousand O*NET-SOC occupations — the essential-skills and transferable-skills tables together (importance and level ratings) plus the technology-software table (with hot-technology and in-demand flags). Because every record carries the onet_soc_code, associations join cleanly to any occupation-keyed dataset, and element_name lets you pivot the other way — from a skill or a technology to every occupation that uses it. Pull one occupation family with an onetSocCodePrefix, one technology with elementContains, or the whole table to seed a skills graph.

Working with the associations

The two association types answer different questions and share one key. Skill rows tell you how important a skill is to an occupation and what level it demands — the IM and LV scale values let you rank skills within a role or compare the same skill across roles. Software rows tell you which technologies a role uses and whether O*NET considers each one a hot or in-demand technology — useful for tooling-adoption analysis and for aligning a candidate's stack to a role. Because both share onet_soc_code, a single occupation's full profile is one filtered pull, and because both carry element_name, the inverse lookup — every occupation that uses a given skill or technology — is equally direct. The hot_technology and in_demand flags travel verbatim so a downstream model can weight emerging tools without re-deriving that signal.

Cost guidance

One association_record per delivered row: cost = rows delivered × your plan's per-record price. Scope with onetSocCodePrefix, associationType, or elementContains to keep runs bounded — one occupation family or one technology is typically a few hundred to a few thousand rows. 1,000 associations is $50 (Free) / $33.50 (Gold); the full ~94,000-row table is a one-time seed you can then keep fresh with small targeted pulls.

Honest limitations

• Reference, not a roster. This is occupational reference data describing work — skills and software by occupation — not a list of people, employers, or job postings.
• As published. Ratings, flags, element names, and titles are exactly as O*NET distributes them; this actor reformats, it does not recompute or re-survey.
• Two association types. Skill rows carry ratings; software rows carry technology examples and flags — the unused fields are null by design, not missing data.
• No person data of any kind.

Differentiation

This is a net-new category for the NexGenData fleet — no existing fleet actor covers ONET occupation-to-skill or occupation-to-software associations. It is distinct from any jobs, hiring, or staffing cell (those carry postings or headcounts, not the standardized skill/technology vocabulary) and from generic occupation directories. Its specific job is the **ONET skills-and-software association table** — the reference mapping that skills-intelligence, talent, and labor-market systems join against.

The NexGen Signal family

Part of the NexGen Signal family of pay-per-record data products:

• US Hospital Ownership Records — CMS Owner Edges
• Higher-Ed Institution Records — US IPEDS
• Grant Opportunity Records — US Grants.gov
• Mine Operator & Safety Records — US MSHA
• Airport Facility Records — US FAA NASR

Source: ONET database (db_30_3), U.S. Department of Labor, Employment and Training Administration (onetcenter.org). ONET data is used under the Creative Commons Attribution 4.0 International (CC BY 4.0) licence — free to reuse, including commercially, with attribution. ONET is a trademark of USDOL/ETA. Occupation-level reference data; no person data of any kind.*