==================================================
 STORYGEN AI — STORY EXPORT
==================================================
Story ID:           17
Title:              AGENTIC TALENT AND SKILLS GAP MAPPING ANALYSIS v4
Owner:              Idhawati MK - Head of People Operations <idha@educlaas.com>
App Type:           Agentic AI App
Input Type:         text
Status:             features_generated
Created:            2026-04-26 14:43:59 UTC
Updated:            2026-04-29 09:41:11 UTC
Features Generated: 2026-04-26 14:46:48 UTC
Total Clusters:     9
Total Features:     17

--- ORIGINAL INPUT ---
Skills Gap Analysis & Personalised Development Platform
Overview
Organisations struggle to objectively assess whether their talents possess the skills required for specific roles, and individuals often lack clarity on how to close their development gaps. This platform creates a unified system that profiles  talent capabilities, maps them against job-specific competency frameworks, visualises skill gaps, and delivers personalised learning paths to help talents reach target roles.
The platform also includes a Skills Passport, a continuously updated profile that captures all talent achievements from training, certifications, and real project experience, providing a dynamic and evidence-based view of capabilities.
Actors
•	Talent: Review their skill gaps, and follows personalised learning paths to qualify for new or advanced roles. 
•	HR : Create profile and defines standardised competency frameworks and aligns them with industry standards like SkillsFuture. 
•	Business Lead: Monitors team members' learning progress and adjusts workloads to support development.
•	AI Agents (Profile, Gap Analysis, Learning, Skills Passport Agent): Automate skill extraction, gap analysis, recommendations, and continuous profile updates. 
•	ClaaS Mentor: Delivers personalised learning, answers questions, evaluates assignments, and adapts learning based on performance.
•	ClaaS Developer: Designs job-role-based curriculum and learning content aligned to competency frameworks.
•	CLaaS Manager: Triggers enrollment and orchestrates execution of personalised learning plans generated by the platform.
Goals
•	Provide a single source of truth for skills data across candidates and roles. 
•	Quantify skill gaps using consistent, role-based benchmarks. 
•	Accelerate workforce readiness by recommending targeted, trackable learning interventions. 
•	Support internal mobility (vertical and lateral moves) through transparent career pathing. 
•	Reduce time spent by HR and candidates on manual skills assessment and training discovery.
•	Maintain a real-time, evidence-based Skills Passport that reflects actual capabilities from learning, certifications, and project work. 
•	Enable learn-to-work capability, where learning is directly embedded into real job tasks and projects.
• Provide adaptive learning journeys that automatically adjust based on learner progress and performance.

User Story
As an Employer, or Organisation, I want to build a digital skills profile, view how it compares to the requirements of the job target role, and receive that talent (candidate) personalised learning path, so that talent can close skill gaps efficiently and qualify for promotions or new opportunities and maintain a continuously updated Skills Passport that reflects verified achievements and real-world capabilities.

Detailed Workflow
1. Profile Creation & Job Mapping
1.	HR uploads a resume; the system uses NLP-based resume parsing to automatically extract skills, experiences, and certifications. 
2.	HR reviews extracted skills and manually adds, edits, or confirms entries to complete their baseline profile. 
3.	HR defines a competency framework for each job role (technical skills, soft skills, required proficiency levels), optionally aligned with industry standards such as SkillsFuture, and integrates external data sources such as job descriptions and succession plans where available.
4.	The system automatically ranks candidates based on their natural fit score against the mapped job requirements. 
2. Skills Gap Identification
6.	HR selects a target role and views a visual comparison (e.g., spider/radar chart) of their current skills versus required skills. 
7.	AI scores the talent's proficiency on a 1–5 scale for each relevant skill, feeding quantitative data into the gap analysis. 
8.	The system automatically detects and highlights "hidden" gaps — skills the candidate did not self-identify but are required for the target role. 
9.	A real-time dashboard visualises the "chasm" between current and target skill levels at both individual and team levels. 
3. Personalised Learning Paths
10.	Based on identified gaps, the system recommends curated learning resources (courses, certifications, and project-based assignments) ranked by relevance. 
11.	AI Mentor delivers personalised learning, provides real-time guidance, and answers learner queries during the learning process. 
12.	The system assigns project-based tasks and automatically evaluates submissions using AI, with optional validation by business leads or managers
13.	If gaps remain, the system automatically generates adaptive learning content. 
14.	Talents views a step-by-step development roadmap showing the sequence of skills to acquire for their target vertical or lateral role.
15.	The personalised learning plan is passed to CLaaS Manager, which triggers enrolment and coordinates execution of the learning journey.
16.	Business Leads monitors team progress in real time through a dashboard and adjusts workloads or provides coaching as needed. 
17.	As the talents complete tasks and projects, their proficiency scores and profiles are updated, and the gap visualisation refreshes automatically.
4. Skills Passport 
18.	The system automatically updates the Skills Passport when learning activities and tasks are completed via Adaptive CLaaS. 
19.	Certifications (internal and external) are added and mapped to relevant skills. 
20.	Project and work experience contributions are recorded, including skills applied and validated by managers, or AI-based assessment. 
21.	Skill proficiency levels are updated based on verified evidence (training, certifications, projects, and AI/peer validation results). 
22.	The Skills Passport provides a real-time, consolidated view of all achievements and capabilities. 
23.	Skills Passport is continuously updated in real time as learning, assessments, and work activities are completed.

Acceptance Criteria
•	HR can upload a resume (PDF/DOCX) and have at least 90% of listed skills automatically extracted via NLP. 
•	HR can manually add, edit, or remove skills after parsing. 
•	HR can create, edit, and publish competency frameworks for each job role, with each competency assigned a defined proficiency scale (1–5). 
•	Competency frameworks can be imported from or aligned with national standards framework
•	The gap analysis view displays a visual chart (e.g., a spider chart) comparing current and required proficiency for each mapped competency. 
•	Submit 1–5 proficiency scores per skill, and scores are timestamped and auditable. 
•	The system automatically flags at least one "hidden gap" when a required skill is absent from the candidate's profile. 
•	Learning recommendations are generated for every identified gap, with at least three options per skill where available. 
•	A development roadmap is presented in sequential steps, distinguishing vertical vs. lateral career paths. 
•	The platform integrates with Adaptive CLaaS via CLaaS Manager to trigger enrolment and execution of personalised learning plans.
•	Business Leads can view a team-level dashboard showing each member's learning progress, completion percentage, and remaining gaps. 
•	All dashboards update in real time (or within a defined SLA, e.g., under 5 minutes of a data change). 
•	Skills Passport automatically updates when learning activities and tasks are completed
•	Certifications are recorded and mapped to skills. 
•	Project-based skill evidence can be added and validated. 
•	Skill levels are updated dynamically based on verified achievements. 
•	AI Mentor can provide real-time learning guidance and answer queries.
•	Project-based assignments can be automatically assessed using AI.
•	Adaptive learning pathways are generated when talent do not meet required proficiency.
•	The platform integrates with external HR systems  to ingest job roles, competencies, and performance data.
•	Skills Passport updates in real time based on learning, project work, and validation inputs.
Assumptions & Constraints
•	Adaptive CLaaS platform and CLaaS Manager are available to execute personalised learning plans.
•	The organisation has (or is willing to define) standardised competency frameworks per role. 
•	Resume parsing relies on NLP and may require human validation for edge cases or non-standard resume formats. 
•	Integration with SkillsFuture or similar frameworks requires access to their skill taxonomies. 
•	Proficiency scoring assumes are available; self-assessment may be used as a supplementary input. 
•	The solution must comply with data privacy regulations (e.g., PDPA, GDPR) given that it stores personal career and performance data. 
•	Initial release supports internal mobility use cases; external recruiting workflows may be added in later phases. 
•	Validation of project-based skills may require approval from business lead.
•	External certification integration may depend on third-party APIs or manual upload. 
•	Integration with HR systems (depends on API availability and data quality.
•	AI-based assessment accuracy may require human validation in critical roles or high-stakes evaluations.

•	Project-based learning requires structured task definitions aligned with job roles.

--- USER STORY ---
# Skills Gap Analysis & Personalised Development Platform

## Overview
This is an **Agentic AI App** that addresses the challenge of objectively assessing talent capabilities against role requirements and closing development gaps in a personalised, evidence-based way. A team of autonomous AI agents collaborates to profile talent, quantify skill gaps, generate adaptive learning journeys, and continuously maintain a real-time Skills Passport — with humans (HR, Business Leads, Talents) overseeing key decisions and validations.

## Actors
- **Talent (Employee/Candidate)**: Reviews their skill profile and gaps, follows personalised learning paths, completes project-based assignments, and progresses toward target roles.
- **HR**: Creates baseline talent profiles, defines and publishes role-based competency frameworks aligned with standards like SkillsFuture, and validates parsed data.
- **Business Lead**: Monitors team learning progress via dashboards, validates project-based skill evidence, and adjusts workloads or provides coaching.
- **Profile Agent (AI)**: Autonomously parses resumes via NLP, extracts skills/experiences/certifications, and structures the baseline talent profile.
- **Gap Analysis Agent (AI)**: Compares talent profiles against role competency frameworks, scores proficiency on a 1–5 scale, detects hidden gaps, and produces visual gap analyses.
- **Learning Agent (AI)**: Curates and ranks learning recommendations, generates adaptive content when gaps persist, and sequences development roadmaps.
- **Skills Passport Agent (AI)**: Continuously updates the Skills Passport in real time based on completed learning, certifications, project evidence, and validations.
- **ClaaS Mentor (AI)**: Delivers personalised learning, answers learner queries in real time, evaluates assignments, and adapts content based on performance.
- **ClaaS Developer**: Designs job-role-based curriculum and learning content aligned with competency frameworks.
- **CLaaS Manager**: Triggers enrolment and orchestrates execution of the personalised learning plans generated by the platform.

## Goals
- Provide a single source of truth for skills data across talents and roles.
- Quantify skill gaps using consistent, role-based benchmarks.
- Accelerate workforce readiness through targeted, trackable, AI-driven learning interventions.
- Support internal mobility (vertical and lateral) via transparent career pathing.
- Reduce HR and talent effort spent on manual skills assessment and training discovery.
- Maintain a real-time, evidence-based Skills Passport reflecting verified capabilities.
- Enable learn-to-work capability where learning is embedded in real job tasks.
- Provide adaptive learning journeys that auto-adjust to learner progress.

## User Story
As an **Employer or Organisation**, I want autonomous AI agents to build a digital skills profile for each talent, compare it against target role requirements, generate and orchestrate a personalised learning path, and continuously maintain a verified Skills Passport — so that talents can efficiently close skill gaps, qualify for promotions or new opportunities, and showcase real-world capabilities backed by evidence.

## Detailed Workflow

### 1. Profile Creation & Job Mapping
1. HR uploads a talent's resume (PDF/DOCX). The **Profile Agent** parses it via NLP and autonomously extracts skills, experiences, and certifications.
2. The Profile Agent flags low-confidence extractions and defers them to HR for manual confirmation. HR adds, edits, or confirms entries to finalise the baseline profile.
3. HR defines or imports a competency framework per role (technical skills, soft skills, proficiency levels), optionally aligned with SkillsFuture; the Profile Agent integrates external data such as job descriptions and succession plans where available.
4. The **Gap Analysis Agent** automatically ranks candidates by natural fit score against role requirements.

### 2. Skills Gap Identification
5. HR or Talent selects a target role; the Gap Analysis Agent generates a visual comparison (spider/radar chart) of current vs. required proficiency.
6. The agent scores each relevant skill on a 1–5 scale, timestamps the result, and writes auditable evidence to the profile.
7. The agent autonomously detects "hidden gaps" — required skills missing from the talent's profile — and highlights them.
8. A real-time dashboard visualises the gap "chasm" at individual and team levels.

### 3. Personalised Learning Paths
9. The **Learning Agent** recommends curated learning resources (courses, certifications, project-based assignments) ranked by relevance, producing at least three options per gap where available.
10. The Learning Agent sequences a step-by-step development roadmap, distinguishing vertical vs. lateral career moves.
11. The personalised plan is handed off to **CLaaS Manager**, which triggers enrolment and orchestrates execution.
12. **ClaaS Mentor** delivers learning, answers queries in real time, and evaluates submissions; it escalates ambiguous or high-stakes assessments to a Business Lead for validation.
13. The Mentor assigns project-based tasks and auto-evaluates submissions; Business Leads optionally validate project evidence.
14. If gaps persist after a learning attempt, the Learning Agent autonomously generates **adaptive learning content** and adjusts the roadmap.
15. Business Leads monitor real-time progress dashboards and adjust workloads or coaching.
16. As tasks complete, the Gap Analysis Agent refreshes proficiency scores and the gap visualisation automatically.

### 4. Skills Passport (Continuous Update)
17. The **Skills Passport Agent** automatically updates the Passport whenever learning activities or project tasks are completed via Adaptive CLaaS.
18. Internal and external certifications are ingested and mapped to relevant skills.
19. Project and work experience are recorded, including skills applied, with validation by managers or AI-based assessment.
20. Skill proficiency levels are updated based on verified evidence (training, certifications, projects, AI/peer validation).
21. The Skills Passport provides a real-time, consolidated, evidence-backed view of capabilities.
22. For high-stakes role qualifications, the Passport Agent defers final certification of proficiency to HR or Business Lead approval.

## Acceptance Criteria

### Profile & Framework
- The Profile Agent extracts at least **90%** of skills listed in standard PDF/DOCX resumes via NLP.
- HR can manually add, edit, or remove skills after parsing.
- HR can create, edit, and publish competency frameworks per role, each with a defined 1–5 proficiency scale.
- Frameworks can be imported from or aligned with national standards (e.g., SkillsFuture).

### Gap Analysis
- A visual chart (spider/radar) displays current vs. required proficiency per competency.
- Proficiency scores (1–5) are timestamped and auditable.
- The Gap Analysis Agent flags at least one "hidden gap" when a required skill is absent from the profile.

### Learning & Orchestration
- Learning recommendations are generated for **every** identified gap, with at least three options per skill where available.
- A sequential development roadmap is presented, distinguishing vertical vs. lateral career paths.
- The platform integrates with Adaptive CLaaS via CLaaS Manager to trigger enrolment and execution.
- ClaaS Mentor provides real-time learning guidance and answers queries.
- Project-based assignments can be auto-assessed by AI.
- Adaptive learning pathways are generated when talents do not meet required proficiency.

### Dashboards & Real-Time Updates
- Business Leads can view a team-level dashboard showing each member's progress, completion %, and remaining gaps.
- All dashboards update in real time (or within a defined SLA — under 5 minutes of a data change).

### Skills Passport
- The Skills Passport auto-updates when learning activities and tasks complete.
- Certifications are recorded and mapped to skills.
- Project-based skill evidence can be added and validated.
- Skill levels update dynamically based on verified achievements.
- The Passport updates in real time based on learning, project work, and validation inputs.

### Integrations
- The platform integrates with external HR systems to ingest job roles, competencies, and performance data.

### Agent Guardrails & Escalation
- The Profile Agent must defer to HR when extraction confidence falls below a defined threshold (e.g., < 80%).
- The Gap Analysis Agent must not finalise proficiency scores for high-stakes promotion decisions without human (HR or Business Lead) sign-off.
- ClaaS Mentor must escalate ambiguous assessment outcomes or learner disputes to a Business Lead.
- The Learning Agent must stop generating adaptive content after a configurable number of unsuccessful attempts and escalate to ClaaS Developer for curriculum review.
- The Skills Passport Agent must require human validation before recording project-based skill evidence in regulated or high-stakes roles.
- All autonomous agent decisions affecting a talent's profile must be logged with rationale and remain auditable.
- Agents must comply with data privacy regulations (PDPA, GDPR) and operate only on data the talent or HR has authorised.

## Assumptions & Constraints
- Adaptive CLaaS platform and CLaaS Manager are available to execute personalised learning plans.
- The organisation has (or will define) standardised competency frameworks per role.
- Resume parsing relies on NLP and may require human validation for edge cases or non-standard formats.
- Integration with SkillsFuture or similar frameworks requires access to their skill taxonomies.
- Proficiency scoring may use self-assessment as supplementary input where objective data is unavailable.
- The solution must comply with data privacy regulations (PDPA, GDPR) given storage of personal career and performance data.
- Initial release supports internal mobility; external recruiting workflows may be added later.
- Validation of project-based skills may require Business Lead approval.
- External certification integration depends on third-party APIs or manual upload.
- HR system integration depends on API availability and data quality.
- AI-based assessment accuracy may require human validation in critical or high-stakes roles.
- Project-based learning requires structured task definitions aligned with job roles.

--- FEATURE LIST SUMMARY ---
This solution enables organisations to objectively assess talent capabilities, quantify skill gaps against role benchmarks, and orchestrate personalised, AI-driven learning journeys that culminate in a verified Skills Passport. Primary actors are Talent, HR, Business Lead, ClaaS Developer, and CLaaS Manager, supported by autonomous agents — Profile, Gap Analysis, Learning, Skills Passport, and ClaaS Mentor. The flow runs from master data setup, resume parsing and profile creation, through gap identification, learning orchestration, mentor-led delivery, and continuous Skills Passport updates, ending in dashboards, governance, and agent observability. The Master Data Configuration cluster holds users, roles, competency frameworks, skill taxonomies, and integration configs. Total of 18 features delivering automated profiling, adaptive learning, and evidence-backed Skills Passport.

Feature Clusters & Features:
• Master Data Configuration
  - 1. User, Role & Access Management — Central directory of talents, HR, business leads, ClaaS staff, and the permissions that govern every workflow.
  - 2. Competency Framework & Role Catalog — Library of role-based competency frameworks with 1–5 proficiency scales, aligned to SkillsFuture and reused across gap analysis and learning.
  - 3. Skill Taxonomy & Reference Lists — Master catalog of skills, certifications, career-path types, and lookup values referenced across profiles and learning content.
  - 4. External System Integration Registry — Configuration store for HR system, SkillsFuture, certification providers, and Adaptive CLaaS API connections.
• Talent Profile Creation
  - 5. Resume Upload & NLP Parsing — HR uploads resumes and the Profile Agent autonomously extracts skills, experiences, and certifications into a structured profile.
  - 6. Low-Confidence Review & HR Validation — Flags uncertain extractions for HR confirmation and lets HR add, edit, or remove parsed skills before finalising the baseline profile.
• Role Mapping & Fit Scoring
  - 7. Target Role Selection & Natural Fit Ranking — Talents or HR pick a target role and the Gap Analysis Agent ranks candidates by natural fit against role requirements.
• Skills Gap Analysis
  - 8. Proficiency Scoring & Visual Gap Comparison — Generates timestamped 1–5 scores and spider/radar charts comparing current vs. required proficiency per competency.
  - 9. Hidden Gap Detection & Audit Evidence — Detects required skills missing from the profile, highlights them, and writes auditable rationale for every score.
• Personalised Learning Path
  - 10. Learning Recommendation & Roadmap Sequencing — The Learning Agent curates at least three ranked options per gap and sequences a vertical/lateral development roadmap.
  - 11. Enrolment Orchestration via CLaaS Manager — Hands off the personalised plan to CLaaS Manager which triggers enrolment and orchestrates execution in Adaptive CLaaS.
• Learning Delivery & Assessment
  - 12. ClaaS Mentor Real-Time Guidance — Delivers learning content, answers learner queries instantly, and adapts pacing based on individual performance.
  - 13. Project-Based Assignment & Auto-Evaluation — Assigns real-job tasks, auto-evaluates submissions, and routes ambiguous or high-stakes outcomes to a Business Lead.
  - 14. Adaptive Content Generation on Persistent Gaps — When gaps remain after attempts, the Learning Agent generates new adaptive content and escalates to ClaaS Developer after repeated failures.
• Progress Monitoring
  - 15. Real-Time Team & Talent Dashboards — Business Leads and Talents view live progress, completion %, and remaining gaps with sub-five-minute refresh SLA.
• Skills Passport Maintenance
  - 16. Continuous Skills Passport Updates — The Skills Passport Agent auto-updates verified skills, certifications, and project evidence as activities complete.
  - 17. High-Stakes Validation & Certification Approval — Defers final proficiency certification for regulated or promotion-critical roles to HR or Business Lead sign-off.
• Agent Orchestration & Governance
  - 18. Agent Audit Log, Guardrails & Privacy Compliance — Logs every autonomous decision with rationale, enforces confidence thresholds, escalation rules, and PDPA/GDPR compliance across all agents.

--- FEATURE LIST (17 features across 9 clusters) ---

#2 | Cluster: Master Data Configuration | Feature: Competency Framework & Role Catalog
  Description: Defines the role-based competency frameworks used as the benchmark for every gap analysis. Supports SkillsFuture alignment and versioning.
  Workflow:
    1. HR creates a role with title, family, and level.
    2. Adds required competencies with target proficiency 1–5.
    3. Imports/aligns to SkillsFuture taxonomy.
    4. Reviews and publishes the framework.
    5. Versions and archives outdated frameworks.
  Table:       role_competency_frameworks
  Columns:     id (bigint, pk), role_code (varchar 50), role_title (varchar 200), skill_id (bigint, fk), required_level (smallint), framework_version (varchar 20), is_published (boolean), source_standard (varchar 50), created_at (timestamp)
  Actor:       HR
  AI Agent:    None
  ----
#3 | Cluster: Master Data Configuration | Feature: Skill Taxonomy & Reference Lists
  Description: Master catalog of skills, certifications, and lookup values referenced everywhere. Ensures consistent terminology across profiles, frameworks, and learning content.
  Workflow:
    1. Admin imports baseline skill taxonomy.
    2. Adds custom skills, categories (technical/soft), and synonyms.
    3. Maintains certification catalog and career-path types.
    4. Approves taxonomy changes via review workflow.
    5. Publishes for use across platform.
  Table:       skills_catalog
  Columns:     id (bigint, pk), skill_name (varchar 200), category (varchar 50), parent_skill_id (bigint, fk), synonyms (text), source (varchar 50), is_active (boolean), created_at (timestamp)
  Actor:       HR / Admin
  AI Agent:    None
  ----
#4 | Cluster: Master Data Configuration | Feature: External System Integration Registry
  Description: Stores connection settings and field mappings for all external integrations. Drives data ingestion for roles, performance data, certifications, and learning execution.
  Workflow:
    1. Admin registers external system (HRIS, SkillsFuture, Adaptive CLaaS, certification providers).
    2. Stores API endpoints, credentials, and data mappings.
    3. Tests connection and toggles status.
    4. Schedules sync jobs.
    5. Monitors integration health.
  Table:       integration_configs
  Columns:     id (bigint, pk), system_name (varchar 100), endpoint_url (varchar 500), auth_type (varchar 50), credentials_ref (varchar 200), sync_frequency (varchar 50), status (varchar 20), last_sync_at (timestamp)
  Actor:       System Administrator
  AI Agent:    None
  ----
#5 | Cluster: Talent Profile Creation | Feature: Resume Upload & NLP Parsing
  Description: Autonomously parses resumes to build a structured baseline talent profile. Targets ≥90% extraction accuracy on standard formats.
  Workflow:
    1. HR uploads resume (PDF/DOCX) for a talent.
    2. Profile Agent runs NLP extraction for skills, experience, certifications.
    3. Agent normalises terms via skill taxonomy.
    4. Agent creates baseline talent profile record.
    5. Logs extraction confidence and audit trail.
  Table:       talent_profiles
  Columns:     id (bigint, pk), talent_user_id (bigint, fk), resume_file_url (varchar 500), parsed_payload (jsonb), extraction_confidence (numeric), parsing_status (varchar 30), created_by (bigint, fk), created_at (timestamp)
  Actor:       HR / Profile Agent
  AI Agent:    Profile Agent
  ----
#6 | Cluster: Talent Profile Creation | Feature: Low-Confidence Review & HR Validation
  Description: Human-in-the-loop validation for uncertain NLP extractions. Ensures profile quality before downstream gap analysis.
  Workflow:
    1. Profile Agent flags entries with confidence below threshold (e.g., <80%).
    2. Items routed to HR review queue.
    3. HR confirms, edits, or removes each entry.
    4. Talent may also self-confirm assigned items.
    5. Profile finalised and locked as baseline.
  Table:       profile_skill_validations
  Columns:     id (bigint, pk), profile_id (bigint, fk), skill_id (bigint, fk), extracted_value (varchar 200), confidence (numeric), validation_status (varchar 30), validated_by (bigint, fk), validated_at (timestamp)
  Actor:       HR
  AI Agent:    Profile Agent
  ----
#7 | Cluster: Role Mapping & Fit Scoring | Feature: Target Role Selection & Natural Fit Ranking
  Description: Computes a natural-fit score between a talent's profile and target role requirements. Powers role recommendations and candidate shortlisting.
  Workflow:
    1. Talent or HR selects target role from catalog.
    2. Gap Analysis Agent loads role framework.
    3. Agent computes natural fit score against profile.
    4. Ranks candidates for the role.
    5. Persists ranking with timestamp.
  Table:       role_fit_scores
  Columns:     id (bigint, pk), profile_id (bigint, fk), role_code (varchar 50), fit_score (numeric), rank_position (int), framework_version (varchar 20), computed_at (timestamp)
  Actor:       Talent / HR
  AI Agent:    Gap Analysis Agent
  ----
#8 | Cluster: Skills Gap Analysis | Feature: Proficiency Scoring & Visual Gap Comparison
  Description: Produces auditable proficiency scores and visual gap charts per competency. Enables consistent, role-based benchmarking.
  Workflow:
    1. Gap Analysis Agent scores each role-required skill 1–5.
    2. Compares current vs. required proficiency.
    3. Generates spider/radar visualisation.
    4. Stores timestamped scores with evidence.
    5. Publishes to talent and HR dashboards.
  Table:       skill_gap_assessments
  Columns:     id (bigint, pk), profile_id (bigint, fk), role_code (varchar 50), skill_id (bigint, fk), current_level (smallint), required_level (smallint), evidence_ref (text), scored_at (timestamp)
  Actor:       Gap Analysis Agent
  AI Agent:    Gap Analysis Agent
  ----
#9 | Cluster: Skills Gap Analysis | Feature: Hidden Gap Detection & Audit Evidence
  Description: Detects required skills entirely absent from the talent's profile and logs evidence. Ensures coverage of non-obvious gaps in addition to scored ones.
  Workflow:
    1. Agent scans required skills missing from profile.
    2. Flags as hidden gaps with rationale.
    3. Writes audit log entry with reasoning.
    4. Surfaces hidden gaps in chart and dashboard.
    5. Triggers learning recommendations downstream.
  Table:       hidden_gap_findings
  Columns:     id (bigint, pk), assessment_id (bigint, fk), skill_id (bigint, fk), rationale (text), detected_at (timestamp), agent_run_id (varchar 100)
  Actor:       Gap Analysis Agent
  AI Agent:    Gap Analysis Agent
  ----
#10 | Cluster: Personalised Learning Path | Feature: Learning Recommendation & Roadmap Sequencing
  Description: Generates ranked learning options per gap and a sequenced development roadmap. Distinguishes vertical vs. lateral career moves.
  Workflow:
    1. Learning Agent ingests gap list per talent.
    2. Curates ≥3 ranked options per gap (course, cert, project).
    3. Sequences into vertical or lateral roadmap.
    4. Estimates duration and effort.
    5. Publishes draft roadmap for talent review.
  Table:       learning_roadmaps
  Columns:     id (bigint, pk), profile_id (bigint, fk), target_role_code (varchar 50), path_type (varchar 20), step_sequence (jsonb), estimated_duration_days (int), status (varchar 30), created_at (timestamp)
  Actor:       Talent / Learning Agent
  AI Agent:    Learning Agent
  ----
#11 | Cluster: Personalised Learning Path | Feature: Enrolment Orchestration via CLaaS Manager
  Description: Hands off the personalised plan to CLaaS Manager for execution. Coordinates enrolment and scheduling on the Adaptive CLaaS platform.
  Workflow:
    1. Talent or HR confirms roadmap.
    2. CLaaS Manager receives plan via API.
    3. Triggers enrolment in Adaptive CLaaS.
    4. Schedules learning steps and notifications.
    5. Reports enrolment status back to platform.
  Table:       learning_enrolments
  Columns:     id (bigint, pk), roadmap_id (bigint, fk), step_id (varchar 100), claas_course_id (varchar 100), enrolment_status (varchar 30), enrolled_at (timestamp), completed_at (timestamp)
  Actor:       CLaaS Manager
  AI Agent:    Learning Agent
  ----
#12 | Cluster: Learning Delivery & Assessment | Feature: ClaaS Mentor Real-Time Guidance
  Description: AI mentor delivers learning content and answers learner questions in real time. Adapts difficulty and pacing based on performance.
  Workflow:
    1. Talent enters learning module.
    2. ClaaS Mentor delivers personalised content.
    3. Answers queries via conversational interface.
    4. Adapts pacing to learner performance.
    5. Logs interactions and outcomes.
  Table:       mentor_interactions
  Columns:     id (bigint, pk), enrolment_id (bigint, fk), interaction_type (varchar 50), prompt_text (text), response_text (text), performance_signal (numeric), occurred_at (timestamp)
  Actor:       Talent
  AI Agent:    ClaaS Mentor
  ----
#13 | Cluster: Learning Delivery & Assessment | Feature: Project-Based Assignment & Auto-Evaluation
  Description: Embeds learning in real job tasks via project assignments with AI auto-grading. Escalates ambiguous outcomes to Business Lead validation.
  Workflow:
    1. Mentor assigns project task aligned with target skill.
    2. Talent submits deliverable.
    3. Mentor auto-evaluates submission.
    4. Ambiguous/high-stakes results escalated to Business Lead.
    5. Result captured as skill evidence.
  Table:       project_assignments
  Columns:     id (bigint, pk), enrolment_id (bigint, fk), task_definition (text), submission_url (varchar 500), ai_score (numeric), ai_feedback (text), validation_status (varchar 30), validator_id (bigint, fk), evaluated_at (timestamp)
  Actor:       Talent / Business Lead
  AI Agent:    ClaaS Mentor
  ----
#14 | Cluster: Learning Delivery & Assessment | Feature: Adaptive Content Generation on Persistent Gaps
  Description: Auto-creates remedial content when gaps persist and reshapes the roadmap. Halts and escalates to curriculum design after repeated failures.
  Workflow:
    1. Learning Agent detects unmet proficiency after attempt.
    2. Generates new adaptive content tailored to weak areas.
    3. Updates roadmap and re-enrols learner.
    4. After N failed attempts, escalates to ClaaS Developer.
    5. Logs every adaptation decision.
  Table:       adaptive_content_events
  Columns:     id (bigint, pk), roadmap_id (bigint, fk), skill_id (bigint, fk), attempt_count (int), generated_content_ref (varchar 500), escalation_status (varchar 30), created_at (timestamp)
  Actor:       Learning Agent / ClaaS Developer
  AI Agent:    Learning Agent
  ----
#15 | Cluster: Progress Monitoring | Feature: Real-Time Team & Talent Dashboards
  Description: Live dashboards for talents and Business Leads showing progress, completion, and remaining gaps. Supports interventions and coaching decisions.
  Workflow:
    1. System aggregates progress, completion %, gaps in real time.
    2. Talent views personal dashboard.
    3. Business Lead views team-level view.
    4. SLA enforces refresh under 5 minutes.
    5. Lead can adjust workload or coach directly from dashboard.
  Table:       progress_dashboard_snapshots
  Columns:     id (bigint, pk), talent_user_id (bigint, fk), team_id (bigint, fk), completion_pct (numeric), open_gaps (int), last_activity_at (timestamp), refreshed_at (timestamp)
  Actor:       Talent / Business Lead
  AI Agent:    Gap Analysis Agent
  ----
#16 | Cluster: Skills Passport Maintenance | Feature: Continuous Skills Passport Updates
  Description: Continuously refreshes the Skills Passport from completed learning, certifications, and project evidence. Provides a real-time, evidence-backed view of capabilities.
  Workflow:
    1. Skills Passport Agent listens to learning, project, certification events.
    2. Maps activities to skills via taxonomy.
    3. Updates proficiency levels with verified evidence.
    4. Records timestamp and source of each update.
    5. Publishes refreshed Passport view in real time.
  Table:       skills_passport_entries
  Columns:     id (bigint, pk), talent_user_id (bigint, fk), skill_id (bigint, fk), proficiency_level (smallint), evidence_type (varchar 50), evidence_ref (varchar 500), verified (boolean), updated_at (timestamp)
  Actor:       Talent
  AI Agent:    Skills Passport Agent
  ----
#17 | Cluster: Skills Passport Maintenance | Feature: High-Stakes Validation & Certification Approval
  Description: Requires human sign-off before recording proficiency for regulated or promotion-critical skills. Preserves trust and compliance in the Passport.
  Workflow:
    1. Passport Agent identifies entry tied to high-stakes/regulated role.
    2. Routes to HR or Business Lead for sign-off.
    3. Reviewer validates evidence and approves or rejects.
    4. Approval recorded with rationale.
    5. Passport updated only on approval.
  Table:       passport_approvals
  Columns:     id (bigint, pk), passport_entry_id (bigint, fk), approver_id (bigint, fk), approval_status (varchar 20), rationale (text), high_stakes_flag (boolean), decided_at (timestamp)
  Actor:       HR / Business Lead
  AI Agent:    Skills Passport Agent
  ----
#18 | Cluster: Agent Orchestration & Governance | Feature: Agent Audit Log, Guardrails & Privacy Compliance
  Description: Centralised audit, guardrail, and privacy framework for every AI agent decision. Ensures explainability, escalation, and regulatory compliance.
  Workflow:
    1. Every agent action logged with input, output, rationale.
    2. Confidence thresholds enforced for autonomous actions.
    3. Escalation rules trigger HR/Business Lead/ClaaS Developer review.
    4. PDPA/GDPR consent checked before data access.
    5. Auditor reviews logs and runs compliance reports.
  Table:       agent_audit_log
  Columns:     id (bigint, pk), agent_name (varchar 100), action_type (varchar 100), subject_user_id (bigint, fk), input_payload (jsonb), output_payload (jsonb), confidence (numeric), escalation_target (varchar 100), consent_basis (varchar 100), created_at (timestamp)
  Actor:       Compliance Officer / System
  AI Agent:    Profile Agent / Gap Analysis Agent / Learning Agent / Skills Passport Agent / ClaaS Mentor
  ----

==================================================
 END OF STORY EXPORT
==================================================
