AI for assessmentand management development.
32 skills, understood through their relationships. Three AI agents turn assessments into descriptions, training priorities and personalised development suggestions.
Understand skills through their relationships
Turning assessment results into useful descriptions takes more than reading individual scores. Assessors need to connect observed skills, identify strengths and development areas, and adapt the report to the person’s journey. We built a multi-agent system that supports this work using assessment data and a relationship matrix spanning 32 skills.
- Multi-agent architecture
- Skill relationships
- Personalised reports
What we designed.
AI supports assessors with report writing: it connects observations and skill relationships, produces coherent descriptions and informs development suggestions. The matrix can be updated without redesigning the whole architecture.
Connect the skills
A matrix identifies which skills to consider together. The system reads overall scores and scores from individual assessment activities in the context of the relationships defined by the model.
Separate description and suggestion
A first agent develops skill descriptions; a second uses these to formulate development suggestions. Each stage has a specific role and an output the next stage can use.
Adapt the content to the report
A third agent adapts the shared material into four outputs for fast assessment, executive assessment and development centres. Assessor feedback guided the evolution of the content.
Descriptions, scores and relationship matrix
Description, suggestion and report agents
Four outputs for assessment and development
The data that gives the assessment context
Quantitative assessments feed the system. The report provides a qualitative reading of the observed skills.
Three elements to read together
32 connected skills
A matrix makes relationships between skills explicit, from problem solving to leadership.
- Strategic thinking
- Communication
- Change management
Assessment evidence
Skill descriptions and overall scores, together with scores from individual observed assessment activities.
- Selected skills
- Overall scores
- Assessment activities
Content for assessors
Descriptions and suggestions aligned with the language and format required by each report.
- Fast assessment
- Executive assessment
- Development centre
Three agents to build the report
The assessor selects skills and enters evaluations. The system develops the content step by step.
Description agent
Develops descriptions from skills, scores and relationships in the matrix.
- Overall scores
- Observed activities
- Skill relationships
Descriptions enriched by context.
Suggestion agent
Uses the first agent’s descriptions to develop suggestions connected to the observed skills.
- Descriptions
- Related skills
- Development
Personalised suggestions with a rationale.
Report agent
Adapts the previous outputs into the four required formats. Text refers to the skills observed, uses the past tense and omits numerical scores.
- Fast assessment
- Executive assessment
- Development centre
Content aligned with the assessment journey.
Four outputs for three assessment journeys
The third agent adapts descriptions and suggestions to the required content.
Skill description
A concise description of observed skills and relevant relationships, within ten lines.
- Journey
- Fast assessment
- Output
- General observations
Training priorities
A reading of effective skills and development areas, considering connections with other skills.
- Journey
- Executive assessment
- Output
- Development areas and rationale
Skill introduction
An overview of the assessed skills and what emerged during the assessment.
- Journey
- Development centre
- Output
- Assessment context
Development suggestions
Personalised suggestions built on the descriptions, accompanied by the reason for each suggestion.
- Journey
- Development centre
- Output
- Actions for the development journey
From the assessors’ model to an adaptable system
The matrix combines the existing skills model with AI-assisted analysis of new relationships. The interactive map shows the links in the project file: selecting a skill highlights its associated capabilities. Node positions make the network readable; links represent relationships defined by the model, not statistical coefficients or cause-and-effect relationships.
A more useful perspective for development professionals
- Which skills should be considered together?
- Which training priorities emerge from the relationships?
- How can the same assessment support different reports?
What it makespossible.
- An integrated view of capabilities through a 32-skill matrix.
- Three agents with distinct roles in content generation.
- Four outputs tailored to different assessment journeys.
- A test platform for gathering assessor feedback.
Want to turn your assessments into development journeys?
Want to turn your assessments into development journeys?
Tell us about your project
AI assistant for job search.
From the first question to the next opportunity. A WhatsApp conversation builds a profile, suggests relevant jobs and keeps candidates engaged.