Do we have clarity on what information drives course recommendations? I know skill gaps play a role, but does it also look at locations, education, job titles, job descriptions, etc.?
Solved
What employee content drives course recommendations?
Best answer by dkreiger
Hi
Course recommendations are primarily skills-driven, but not exclusively skill-gap driven. Eightfold considers the employee’s profile skills, proficiency, skill gaps, skill goals, career/role interests, current role and experience, plus organizational context such as seniority, job family/function, business unit, job code, and location. It also uses course metadata — including skills, difficulty, language, target audience, and inferred metadata from course titles/descriptions — and filters out completed courses. Education may contribute to the employee profile context, but the clearest documented primary signals are skills/proficiency, role/career context, and organizational/course metadata.
What drives course recommendations
The current model looks at:
- Employee skills and skill proficiency — current skills, profile skills, self/manager ratings, or seniority-based fallback where proficiency is missing
- Skill gaps — where an employee’s proficiency is below what their role or target role requires
- Skill goals / learning goals — skills the employee explicitly marks as goals
- Role goals / career interests — courses aligned to skills needed for aspired future roles
- Current role and prior experience — the system considers current role, prior job experience, and existing skills
- Seniority / difficulty alignment — course difficulty is matched to the employee’s current proficiency or seniority context, so learners are not shown only overly basic or overly advanced content
- Job function, business unit, job family, and job code — used as organizational context / target-audience matching so courses are better aligned to the employee’s real work context
- Location — included in the improved model’s employee-attribute matching/calibration
- Course metadata — skills covered, difficulty, expected proficiency, course type, duration, language, target audience, and metadata inferred from course title/description when explicit metadata is incomplete
- Course attendance / completion data — completed courses are filtered out; historical attendance can help calibrate which types of employees a course is relevant for
- Language preference — in Career Navigator learning recommendations, preferred-language courses are surfaced first, with English as fallback
On the specific attributes you asked about
| Attribute | Used? | Notes |
|---|---|---|
| Skill gaps | Yes | Core signal; especially for Skill Gaps feed and Career Navigator course suggestions |
| Location | Yes | Used in improved recommendation calibration / matching |
| Education | Some evidence, but not emphasized in latest docs | A discovery transcript mentions education among profile attributes considered, but the latest product/admin docs emphasize role, experience, skills, proficiency, interests, and org context more strongly |
| Job title / current role | Yes | Current role/title and career-interest roles influence recommendations |
| Job descriptions | Indirectly | The docs don’t frame raw job descriptions as a primary direct signal for course recs. Instead, role skills/role context from Talent Design/job architecture and course title/description-derived skills are used |
| Job family/function / BU / job code | Yes | Explicitly used for organizational context and course calibration |
| Course descriptions | Yes | AI can infer course skills and difficulty from course titles/descriptions when metadata |
- Community Manager
- Answer
Hi
Course recommendations are primarily skills-driven, but not exclusively skill-gap driven. Eightfold considers the employee’s profile skills, proficiency, skill gaps, skill goals, career/role interests, current role and experience, plus organizational context such as seniority, job family/function, business unit, job code, and location. It also uses course metadata — including skills, difficulty, language, target audience, and inferred metadata from course titles/descriptions — and filters out completed courses. Education may contribute to the employee profile context, but the clearest documented primary signals are skills/proficiency, role/career context, and organizational/course metadata.
What drives course recommendations
The current model looks at:
- Employee skills and skill proficiency — current skills, profile skills, self/manager ratings, or seniority-based fallback where proficiency is missing
- Skill gaps — where an employee’s proficiency is below what their role or target role requires
- Skill goals / learning goals — skills the employee explicitly marks as goals
- Role goals / career interests — courses aligned to skills needed for aspired future roles
- Current role and prior experience — the system considers current role, prior job experience, and existing skills
- Seniority / difficulty alignment — course difficulty is matched to the employee’s current proficiency or seniority context, so learners are not shown only overly basic or overly advanced content
- Job function, business unit, job family, and job code — used as organizational context / target-audience matching so courses are better aligned to the employee’s real work context
- Location — included in the improved model’s employee-attribute matching/calibration
- Course metadata — skills covered, difficulty, expected proficiency, course type, duration, language, target audience, and metadata inferred from course title/description when explicit metadata is incomplete
- Course attendance / completion data — completed courses are filtered out; historical attendance can help calibrate which types of employees a course is relevant for
- Language preference — in Career Navigator learning recommendations, preferred-language courses are surfaced first, with English as fallback
On the specific attributes you asked about
| Attribute | Used? | Notes |
|---|---|---|
| Skill gaps | Yes | Core signal; especially for Skill Gaps feed and Career Navigator course suggestions |
| Location | Yes | Used in improved recommendation calibration / matching |
| Education | Some evidence, but not emphasized in latest docs | A discovery transcript mentions education among profile attributes considered, but the latest product/admin docs emphasize role, experience, skills, proficiency, interests, and org context more strongly |
| Job title / current role | Yes | Current role/title and career-interest roles influence recommendations |
| Job descriptions | Indirectly | The docs don’t frame raw job descriptions as a primary direct signal for course recs. Instead, role skills/role context from Talent Design/job architecture and course title/description-derived skills are used |
| Job family/function / BU / job code | Yes | Explicitly used for organizational context and course calibration |
| Course descriptions | Yes | AI can infer course skills and difficulty from course titles/descriptions when metadata |
Sign up
Already have an account? Login
Eightfold SSO - Customers & Employees
Eightfold Customers & Employees: Login/Register HereNOT A CUSTOMER? FILL OUT THE FIELDS BELOW:
Login to the Eightfold Pathfinder Community
No account yet? Create an account
Eightfold SSO - Customers & Employees
Eightfold Customers & Employees: Login/Register HereNOT A CUSTOMER? FILL OUT THE FIELDS BELOW:
Enter your E-mail address. We'll send you an e-mail with instructions to reset your password.