One calm learning companion
Short instructions, appropriate language, patient support and one useful next step at a time.
Technology and architecture
More than one chatbot.
A controlled learning system.
KIDNOVA is conceived as a group of specialised components working together. One part manages consent. Another checks safety. Another chooses the teaching approach. Human beings remain responsible for consequential decisions.
Specialised componentsVerified calculationsHuman authority
ChildParentTeacherSpecialist
↓
Identity, consent and permissions
Child safety gateway
Learning orchestrator
Educational digital twin
Approved knowledge and specialist models
Audit, human oversight and governance
Concept architecture — not completed production software
The governing principle
The child sees simplicity. The system carries the complexity.
A child should not have to understand artificial intelligence, model selection or security controls. Those systems must work quietly behind a clear, age-appropriate learning experience.
Many controlled responsibilities
Consent, permissions, safety, curriculum, adaptation, verification, privacy, audit and human review.
Six understandable layers
Each part has a different job.
The architecture is easier to trust when every major responsibility has a clear home and no single model is allowed to control everything.
Experience layer
What children and adults use
The child experience, parent controls, teacher tools, authorised specialist access and administrative support.
Identity and trust layer
Who may enter, see and decide
Parent consent, school authorisation, age-aware access, role permissions and regional requirements.
Safety layer
What may reach the child
Inputs and outputs are examined for harmful content, personal-information risk, unsafe requests and safeguarding concerns.
Teaching layer
How the next learning step is chosen
The learning orchestrator selects the objective, difficulty, teaching method, support level, feedback and review timing.
Learning intelligence layer
What the system remembers for learning
The educational digital twin maintains mastery, accessibility needs, curriculum pathways and instructional preferences.
Governance layer
How the system remains accountable
Important adaptations are recorded, human authority is preserved and consequential decisions remain reviewable.
Experience layer
Different people need different views.
A child should see only what helps them learn. A parent should see controls and useful progress. A teacher should see classroom patterns. A specialist should see only authorised information relevant to their role.
C
Child
Today’s mission, learning activities, progress and a clear trusted-adult route.
P
Parent
Consent, privacy settings, progress over time, voice controls, download and deletion.
T
Teacher
Curriculum links, skill patterns, assigned activities and useful next-action summaries.
S
Specialist
Permission-controlled educational information that supports—but never overrides—professional judgement.
Learning orchestrator
The system must decide how to teach—not merely what to say.
The learning orchestrator brings together the lesson goal, the child’s present response, curriculum requirements, safety constraints and previous learning evidence.
Lesson objectiveCurrent responseMastery evidenceAccessibility settingsCurriculum pathwaySafety rules
Controlled decision pointLearning
Orchestrator
Orchestrator
Chooses the next responsible teaching move
DifficultyTeaching methodSupport levelFeedback typeReview timingPause or adult involvement
Educational digital twin
A learning model—not a copy of the child.
The twin is intended to remember only what helps the platform support learning. It may track skill mastery, accessibility needs, curriculum progress and which forms of explanation have helped.
Non-negotiable boundary
It must not become a broad psychological profile, intelligence label, location tracker, advertising record or prediction of the child’s future potential.
Which specific skills are emerging, developing, secure or ready for transfer.
Whether pictures, spoken guidance, examples or models have supported understanding.
Information that is not necessary for the child’s learning should not belong in the twin.
Specialised model stack
Different tasks need different controls.
A friendly conversation, a curriculum decision, a mathematical calculation and a safety judgement are not the same kind of task. KIDNOVA should not ask one unrestricted model to perform all of them.
Child conversation
Produces short, respectful, age-appropriate dialogue.
Curriculum reasoning
Connects questions to standards, learning objectives and prerequisite skills.
Mathematical verification
Uses deterministic calculation and symbolic checking where correctness matters.
Safety
Examines content and blocks, changes or escalates unsafe interactions.
Localization
Adapts language and examples while preserving educational meaning.
Digital twin
Updates mastery estimates and helps select an appropriate next activity.
Deterministic verification
A plausible answer is not enough.
Generated language can sound confident while being wrong. Arithmetic, scoring and other high-stakes calculations should therefore use dependable checking tools rather than rely on generated wording alone.
Step 1Proposed explanation
→Step 2Verified calculation or rule
→Step 3Child-safe response
Audit and governance
Important changes should leave an understandable record.
When KIDNOVA materially changes a teaching path, the system should be able to explain what changed, why it changed, what evidence informed the decision and whether a human reviewed it.
Illustrative adaptation recordAR-0247
- What changed?
- Written explanation changed to a visual tens-and-ones model.
- Why?
- The child added the ones correctly but confused the tens.
- Evidence used
- Current answer, help request and prior visual-support preference.
- Result
- The child explained that three tens and seven ones make 37.
- Human review
- Available to the authorised parent or teacher summary.
Current status
A defined architecture is not the same as a finished platform.
The KIDNOVA concept is developed enough to support product discovery, prototype planning, partner discussions and pilot design. It still requires engineering, testing and independent validation.
Concept and governing boundaries
Audiences, learning model, safeguards, system roles and development direction.
Working product experience
Child interface, parent controls, teacher tools, model orchestration and data flows.
Safety and reliability
Security, red-team testing, bias, accessibility, consent, retention and escalation.
Educational effectiveness
Independent evidence that children genuinely learn, retain and transfer skills.