Weft

A vocabulary for evidence-backed career graphs.

Weft describes a person's career as a graph: the roles they held, what they studied, built and wrote, the credentials and awards they received, and the skills each of those involved. Every role, degree and project is its own node, linked to where it happened and to the skills it used, taught or certified, so a skill can always be traced back to the work behind it.

Weft builds on schema.org for the kinds of things (people, organisations, roles, credentials, creative works, defined terms) and adds the relations that schema.org doesn't have, or only has as plain text. Skills can be matched to shared concepts such as ESCO or Wikidata with SKOS mappings. CV and portfolio sites, talent tools and research datasets can all use it; Graph Resume is the first application.

The name comes from weaving: the warp is the timeline of roles, studies and projects, and the weft is the skills threaded through them.

IRI: https://w3id.org/weft
This version: https://w3id.org/weft/1.0 (version 1.0, published 2026-11-05)
Namespace: https://w3id.org/weft#, prefix weft:
Author: Rishi Rajani
Licence: CC BY 4.0: free to use and adapt, including commercially, with credit. Data you describe with Weft is yours.
Terms: 2 classes, 22 relations, 4 value properties
Source: github.com/rishirajani/weft · Changes · Suggest a change

Turtle JSON-LD RDF/XML N-Triples JSON-LD context SHACL shapes Example

How it fits together

The kinds of things come from schema.org: the person is a schema:Person, each job a schema:OrganizationRole, each degree or certificate a schema:EducationalOccupationalCredential, each project a schema:CreativeWork, each skill a schema:DefinedTerm, grouped into skill areas (schema:DefinedTermSet). Only awards and career breaks needed classes of their own.

The relations are Weft's. schema.org mostly links a person straight to an organisation (schema:worksFor, schema:alumniOf) and lists skills on the person as text. In Weft every job, degree and project is its own node, linked to where it happened and to the skills it used, trained or certified. That's the evidence layer: any skill can be traced back to the work behind it, and a skill nothing points to stands out.

Where a Weft relation is a special case of a schema.org property it says so (rdfs:subPropertyOf, owl:inverseOf), so data stays useful to tools that only know schema.org. Skills, languages and tools can be matched to shared concepts in ESCO or Wikidata with skos:exactMatch and skos:closeMatch.

Using it

In Turtle, declare the prefix and describe the career with schema.org types and Weft relations:

@prefix weft:   <https://w3id.org/weft#> .
@prefix schema: <https://schema.org/> .
@prefix ex:     <https://example.org/jordan#> .

ex:me a schema:Person ;
  schema:name "Jordan Example" ;
  weft:worksAs ex:r-senior .

ex:r-senior a schema:OrganizationRole ;
  schema:name "Senior Data Analyst" ;
  schema:startDate "2023-03" ;
  weft:at ex:o-lumen ;
  weft:used ex:s-sql .

ex:o-lumen a schema:Organization ; schema:name "Lumen Freight" .
ex:s-sql a schema:DefinedTerm ; schema:name "SQL" .

As JSON-LD, the context lets you write the same thing as plain JSON:

{
  "@context": "https://w3id.org/weft/context.jsonld",
  "id": "https://example.org/jordan#me",
  "type": "Person",
  "name": "Jordan Example",
  "worksAs": {
    "id": "https://example.org/jordan#r-senior",
    "type": "OrganizationRole",
    "name": "Senior Data Analyst",
    "startDate": "2023-03",
    "at": {
      "id": "https://example.org/jordan#o-lumen",
      "type": "Organization",
      "name": "Lumen Freight"
    },
    "used": [
      {
        "id": "https://example.org/jordan#s-sql",
        "type": "DefinedTerm",
        "name": "SQL"
      }
    ]
  }
}

Then ask the graph questions, for example in SPARQL:

PREFIX weft:   <https://w3id.org/weft#>
PREFIX schema: <https://schema.org/>

# Which roles, projects and courses back each skill?
SELECT ?skill ?evidence WHERE {
  ?x weft:used|weft:trainedIn|weft:shows ?s .
  ?s schema:name ?skill .
  ?x schema:name ?evidence .
} ORDER BY ?skill

To check data, run tools/check.py from the repository (every relation connects the kinds of things it should) and validate it against the SHACL shapes, which also warn about skills nothing backs up.

All relations at a glance

From	Relation	To
`schema:Person`	held role	`schema:OrganizationRole`
`schema:Person`	worked as	`schema:OrganizationRole`
`schema:Person`	works as	`schema:OrganizationRole`
`schema:Person`	studied	`schema:EducationalOccupationalCredential`
`schema:Person`	holds	`schema:EducationalOccupationalCredential`
`schema:Person`	built	`schema:CreativeWork`
`schema:Person`	wrote	`schema:ScholarlyArticle`
`schema:Person`	speaks	`schema:Language`
`schema:Person`	received	`weft:Award`
`schema:Person`	specialises in	`schema:DefinedTermSet`
`schema:Person`	took	`weft:CareerBreak`
`schema:OrganizationRole` or `schema:EducationalOccupationalCredential`	at	`schema:Organization`
`schema:OrganizationRole`	promoted from	`schema:OrganizationRole`
`schema:EducationalOccupationalCredential`	led to	`schema:OrganizationRole`
`schema:EducationalOccupationalCredential`	run with	`schema:Organization`
`schema:CreativeWork`	part of	`schema:EducationalOccupationalCredential` or `schema:CreativeWork`
`weft:CareerBreak`	included	`schema:OrganizationRole` or `schema:EducationalOccupationalCredential` or `schema:CreativeWork`
`weft:Award`	awarded by	`schema:Organization`
`schema:OrganizationRole` or `schema:CreativeWork` or `weft:CareerBreak` or `weft:Award`	used	`schema:DefinedTerm` or `schema:Language`
`schema:EducationalOccupationalCredential`	trained in	`schema:DefinedTerm`
`schema:EducationalOccupationalCredential`	shows	`schema:DefinedTerm`
`schema:DefinedTerm`	belongs to	`schema:DefinedTermSet`

Classes

Award `weft:Award`

A recognition given to the person, for example by an employer. schema.org only has award as a text property, so this makes awards nodes that can be linked to who gave them.

IRI: https://w3id.org/weft#Award
Subclass of: schema:Thing
Compare: schema:award

Team award for the depot analytics rollout is a Award

Career break `weft:CareerBreak`

A period between roles, with what happened during it. Lets a gap be described honestly instead of left blank.

IRI: https://w3id.org/weft#CareerBreak
Subclass of: schema:Thing

Gap year is a Career break

Relations

Examples are from the example, a made-up person.

held role `weft:heldRole`

The person held this role. Use the more specific worked as (finished) or works as (current).

IRI: https://w3id.org/weft#heldRole
From: schema:Person
To: schema:OrganizationRole
Broader than: weft:workedAs, weft:worksAs
Compare: schema:hasOccupation

Usually stated through weft:workedAs, weft:worksAs.

worked as `weft:workedAs`

The person held this role in the past; the role has an end date.

IRI: https://w3id.org/weft#workedAs
From: schema:Person
To: schema:OrganizationRole
Narrower than: weft:heldRole

Jordan Example worked as Data Analyst

works as `weft:worksAs`

The person holds this role now; the role has no end date.

IRI: https://w3id.org/weft#worksAs
From: schema:Person
To: schema:OrganizationRole
Narrower than: weft:heldRole

Jordan Example works as Senior Data Analyst

studied `weft:studied`

The person studied for this degree or certificate, finished or still in progress. (schema:hasCredential would claim the credential is already held.)

IRI: https://w3id.org/weft#studied
From: schema:Person
To: schema:EducationalOccupationalCredential
Compare: schema:hasCredential, schema:alumniOf

Jordan Example studied Analytics Engineering Certificate
Jordan Example studied B.Sc. Statistics

holds `weft:holds`

The person holds this certification.

IRI: https://w3id.org/weft#holds
From: schema:Person
To: schema:EducationalOccupationalCredential
Narrower than: schema:hasCredential

Jordan Example holds Cloud Data Practitioner

built `weft:built`

The person built or authored this project.

IRI: https://w3id.org/weft#built
From: schema:Person
To: schema:CreativeWork
Compare: schema:creator

Jordan Example built Delivery delay dashboard

wrote `weft:wrote`

The person is an author of this publication.

IRI: https://w3id.org/weft#wrote
From: schema:Person
To: schema:ScholarlyArticle
Inverse of: schema:author

Jordan Example wrote A short note on seasonal parcel forecasts

speaks `weft:speaks`

The person speaks this language; how well is given by proficiency on the language.

IRI: https://w3id.org/weft#speaks
From: schema:Person
To: schema:Language
Narrower than: schema:knowsLanguage

Jordan Example speaks English
Jordan Example speaks German

received `weft:received`

The person received this award.

IRI: https://w3id.org/weft#received
From: schema:Person
To: weft:Award
Compare: schema:award

Jordan Example received Team award for the depot analytics rollout

specialises in `weft:focusOn`

One of the person's skill areas.

IRI: https://w3id.org/weft#focusOn
From: schema:Person
To: schema:DefinedTermSet

Jordan Example specialises in Data engineering
Jordan Example specialises in Statistics

took `weft:took`

The person took this career break.

IRI: https://w3id.org/weft#took
From: schema:Person
To: weft:CareerBreak

Jordan Example took Gap year

at `weft:at`

Where a role was held (the employer) or where a degree or certificate was studied (the school).

IRI: https://w3id.org/weft#at
From: schema:OrganizationRole or schema:EducationalOccupationalCredential
To: schema:Organization
Compare: schema:worksFor, schema:alumniOf

Analytics Engineering Certificate at Example Data Academy
B.Sc. Statistics at Example University of Applied Sciences
Data Analyst at Lumen Freight

promoted from `weft:promotedFrom`

This role followed the earlier role through a promotion at the same organisation.

IRI: https://w3id.org/weft#promotedFrom
From: schema:OrganizationRole
To: schema:OrganizationRole

Senior Data Analyst promoted from Data Analyst

led to `weft:ledTo`

A programme of study that led directly to this role, for example a certificate run with an employer.

IRI: https://w3id.org/weft#ledTo
From: schema:EducationalOccupationalCredential
To: schema:OrganizationRole

Analytics Engineering Certificate led to Data Analyst

run with `weft:runWith`

An organisation that co-ran this programme with the school.

IRI: https://w3id.org/weft#runWith
From: schema:EducationalOccupationalCredential
To: schema:Organization

Analytics Engineering Certificate run with Lumen Freight

part of `weft:partOf`

This project was done as part of a degree or a larger work, for example a thesis within a master's.

IRI: https://w3id.org/weft#partOf
From: schema:CreativeWork
To: schema:EducationalOccupationalCredential or schema:CreativeWork
Compare: schema:isPartOf

Bachelor's thesis: forecasting weekly parcel volumes part of B.Sc. Statistics

included `weft:included`

Something done during a career break, such as freelance work or a certification.

IRI: https://w3id.org/weft#included
From: weft:CareerBreak
To: schema:OrganizationRole or schema:EducationalOccupationalCredential or schema:CreativeWork

Gap year included Cloud Data Practitioner
Gap year included Food bank stock tracker (volunteer)

awarded by `weft:awardedBy`

The organisation that gave the award.

IRI: https://w3id.org/weft#awardedBy
From: weft:Award
To: schema:Organization
Compare: schema:recognizedBy

Team award for the depot analytics rollout awarded by Lumen Freight

used `weft:used`

A skill put to use in this role, project or career break, including a language used at work. Together with trained in and shows, it ties each skill to the work behind it.

IRI: https://w3id.org/weft#used
From: schema:OrganizationRole or schema:CreativeWork or weft:CareerBreak or weft:Award
To: schema:DefinedTerm or schema:Language
Compare: schema:skills

Team award for the depot analytics rollout used dbt
Gap year used Spreadsheets
Delivery delay dashboard used dbt

trained in `weft:trainedIn`

A skill taught or practised during this degree or course.

IRI: https://w3id.org/weft#trainedIn
From: schema:EducationalOccupationalCredential
To: schema:DefinedTerm
Compare: schema:teaches

Analytics Engineering Certificate trained in SQL
B.Sc. Statistics trained in Python
B.Sc. Statistics trained in Statistical modelling

shows `weft:shows`

A skill this certification is evidence of.

IRI: https://w3id.org/weft#shows
From: schema:EducationalOccupationalCredential
To: schema:DefinedTerm
Compare: schema:competencyRequired

Cloud Data Practitioner shows SQL

belongs to `weft:inGroup`

The skill area this skill is grouped under. Weft doesn't limit how many; an application can (Graph Resume allows one).

IRI: https://w3id.org/weft#inGroup
From: schema:DefinedTerm
To: schema:DefinedTermSet
Narrower than: schema:inDefinedTermSet

dbt belongs to Data engineering
Python belongs to Data engineering
Spreadsheets belongs to Data engineering

Value properties

kind `weft:kind`

An application's own label for what kind of item this is (for example Role, Education, Certification, Project, Skill). Several kinds can share a schema.org type: a degree and a certification are both schema:EducationalOccupationalCredential, and kind keeps them apart without new classes.

IRI: https://w3id.org/weft#kind
To: xsd:string

Team award for the depot analytics rollout kind “Award”

highlight `weft:highlight`

One achievement or responsibility in a role, degree or project, as one short statement (like a bullet point on a CV). Order is not significant.

IRI: https://w3id.org/weft#highlight
To: xsd:string

Data Analyst highlight “Replaced a weekly spreadsheet report with a SQL model and a dashboard the operations tea…”

location `weft:location`

Where it took place, as free text (city and country, or remote). Use schema:location for a structured place.

IRI: https://w3id.org/weft#location
To: xsd:string
Compare: schema:location

Data Analyst location “Hamburg, Germany”

proficiency `weft:proficiency`

How well a language is spoken: a CEFR level (A1 to C2) or Native, Fluent or Basic.

IRI: https://w3id.org/weft#proficiency
From: schema:Language
To: xsd:string

English proficiency “Native”

Weft reuses or links to these rather than replacing them.

Used by

Graph Resume: Rishi Rajani's résumé as an interactive knowledge graph, published as Turtle and JSON-LD. Data (Turtle)

Using Weft somewhere? Open an issue to be listed.

Citing

Rishi Rajani (2026). Weft: a vocabulary for evidence-backed career graphs, version 1.0. https://w3id.org/weft/1.0

Generated from weft.ttl. Term IRIs such as https://w3id.org/weft#used open their section on this page.
© 2026 Rishi Rajani. Weft and its documentation are licensed under CC BY 4.0. Data described with Weft belongs to whoever publishes it.

Weft

Award weft:Award

Career break weft:CareerBreak

held role weft:heldRole

worked as weft:workedAs

works as weft:worksAs

studied weft:studied

holds weft:holds

built weft:built

wrote weft:wrote

speaks weft:speaks

received weft:received

specialises in weft:focusOn

took weft:took

at weft:at

promoted from weft:promotedFrom

led to weft:ledTo

run with weft:runWith

part of weft:partOf

included weft:included

awarded by weft:awardedBy

used weft:used

trained in weft:trainedIn

shows weft:shows

belongs to weft:inGroup

kind weft:kind

highlight weft:highlight

location weft:location

proficiency weft:proficiency

ESCO and Wikidata

JSON Resume and HR Open Standards

Commercial skills catalogues