Role Intelligence — JD Decomposition & RoleModel
ES Reference: Phase 2 (INTOS-ES-P2-001) — Role Intelligence Engine, Section 4
Overview
The Role Intelligence Engine is the analytical foundation behind Role Match v2 and Interview Kit 2.0. It decomposes a job description into a structured, versioned RoleModel — extracting business outcomes, requirements, competencies, seniority expectations, likely interview rounds, and knockout risks. Every element is transparent, source-traced, and user-correctable.
Role Intelligence is not an ATS-pass prediction or offer-probability engine. It is a structured analysis tool that helps you understand what a role really demands and how your evidence aligns.
Key Principles
- Every element stores its source — requirement text, confidence level, and model/prompt version
- Users can correct requirements and mark items as irrelevant or missing
- Regeneration only on JD version change or explicit user refresh — opening a kit never regenerates
- Interview rounds are hypotheses, not facts — derived from role type, seniority, and JD signals
- No predictions — no ATS compatibility, no offer probability, no hiring likelihood
The RoleModel Structure
| Component | Description |
|---|---|
| Role Identity | Title, family, seniority band, location/work model |
| Business Outcomes | Expected results, success measures, mandate scope |
| Requirements | Mandatory and preferred, with importance, source, and confidence per requirement |
| Competencies | Expected depth levels and behavioural indicators per competency |
| Leadership & Commercial | P&L scope, team size, strategic accountability, commercial expectations |
| Interview Rounds | Likely round types (recruiter, manager, technical, behavioural, leadership, case, HR/offer), focus areas and confidence |
| Knockout Risks | Mandatory requirements that would disqualify if unmet |
| Research Facts | Public company/role facts with citations, verified-at date, and freshness |
Screens in This Feature Area
1. Role Intelligence View (/dashboard/interview-kits/role-intelligence)
View the decomposed RoleModel for a job, with expandable sections for each component.
Buttons & Actions:
| Button | Behavior |
|---|---|
| Generate / Refresh RoleModel | Start async JD decomposition (requires JD version change or explicit trigger) |
| Expand Component | View detailed requirement list per component |
| Correct Requirement | Mark a requirement as incorrect, adjust importance, or add missing items |
| View Source | See the source JD text and AI confidence for each element |
| Mark as Irrelevant | Remove a low-confidence or incorrect requirement from calculations |
| Add Missing Requirement | Manually add a requirement the AI missed |
| Link to Evidence | Connect a requirement to Career Vault evidence items |
2. RoleModel History
Track changes to the RoleModel over time as JDs are updated and requirements are corrected.
Buttons & Actions:
| Button | Behavior |
|---|---|
| Version Diff | Compare two RoleModel versions side-by-side |
| Restore Version | Revert to a previous RoleModel version |
Requirement Properties
Each requirement in the RoleModel carries:
| Property | Description |
|---|---|
| Type | Mandatory or Preferred |
| Importance | Critical / High / Medium / Nice-to-have |
| Source | Specific JD text or public research citation |
| Confidence | AI confidence level (High / Medium / Low) |
| Status | Confirmed / Corrected / Marked Irrelevant / Missing |
| Linked Evidence | Connected Career Vault evidence items (if any) |
Interview Round Hypotheses
Based on role type, seniority, and JD signals, the engine hypothesizes likely interview rounds:
| Round Type | Typical Focus | Confidence |
|---|---|---|
| Recruiter Screen | Motivation, logistics, compensation expectations | High |
| Hiring Manager | Role fit, experience depth, team contribution | High |
| Technical / Architecture | Domain expertise, system design, problem-solving | Role-dependent |
| Behavioural / Leadership | People management, conflict, culture, decision-making | High |
| Case / Presentation | Analytical thinking, structured communication | Role-dependent |
| Board / Executive Panel | Strategy, P&L, transformation, governance | Senior roles |
| HR / Offer | Compensation, benefits, start date, references | High |
Role/Seniority Taxonomy
The engine uses a reusable taxonomy while preserving company-specific needs:
| Seniority Band | Typical Scope | Commercial Expectations |
|---|---|---|
| Individual Contributor | Task/project ownership | Delivery against goals |
| Senior IC / Lead | Technical/domain leadership, mentoring | Team-level impact |
| Manager | Team management, hiring, performance | Budget/headcount accountability |
| Senior Manager | Multi-team, strategy input, org design | P&L awareness |
| Director | Org design, strategy, budget ownership | P&L responsibility |
| VP / C-level | Function/business leadership, transformation | Revenue, margin, market impact |
Related ES Requirements
| ID | Requirement |
|---|---|
| P2-RI-001 | Create versioned RoleModel from JD, Job data and approved public research |
| P2-RI-002 | Extract outcomes, requirements, competencies, seniority, leadership/commercial expectations and knockout risks |
| P2-RI-003 | Model likely interview rounds and interviewer functions as preparation hypotheses |
| P2-RI-004 | Each element stores source, confidence, importance and model/prompt version |
| P2-RI-005 | Users can correct requirements and mark irrelevant/missing items |
| P2-RI-006 | Regenerate only when JD version changes or user explicitly refreshes |
| P2-RI-007 | Use reusable role/seniority taxonomy while preserving company-specific needs |
Related Docs
- [[role-match]] — Role Match v2 maps each RoleModel requirement to your evidence
- [[interview-kits]] — Kit 2.0 uses the RoleModel for round plans, question banks, and briefs
- [[readiness]] — Readiness Index uses requirement states from Role Match
- [[career-vault]] — Evidence Vault provides the evidence foundation for requirement matching