Humanity
Protect dignity, agency, wellbeing, and inclusion.
Check: A person can understand the next step. A consequential action has a human owner. Accessibility needs are treated as product requirements.
Learn with purpose. Build with accountability.
HAGANOLOGY gives AIOT a practical standard for designing learning, technology, and decisions around people. It helps a learner understand what a feature is for, what it uses, what it cannot know, and when a human should review the next step.
The five-pillar test
Every AIOT experience should pass these questions
The framework is used in course design, admissions, dashboards, campaigns, partner delivery, and platform operations. It is a working quality standard, not decorative language.
AI
Use automation to assist people without disguising uncertainty or replacing accountable judgement.
Check: The data basis is named. Confidence and limitations are visible. AI output is reviewable before high-impact action.
Governance
Make roles, consent, privacy, audit, retention, and escalation explicit.
Check: The source of record is known. Sensitive fields are minimised. Exceptions and approvals are recorded.
Alignment
Keep every feature connected to an approved purpose, context, and measurable outcome.
Check: The user goal is stated plainly. The system does not optimise an unrelated proxy. The next action matches the programme or service purpose.
Nurture
Turn signals into support, learning, feedback, and a better next attempt.
Check: A stalled user receives a humane route to help. Progress can resume after interruption. Feedback informs improvement without surveillance.
Explainability rail
Understand the next step before you take it
AIOT records a concise, privacy-minimised explanation for selected data-driven actions. The record includes purpose, signals, limits, confidence, human review, and a clear next action.
Example learner explanation
- Purpose
- Help a learner choose a safe, accessible, and useful next step.
- Information used
- programme interest, learning mode, accessibility preference
- Not used
- No hidden sensitive profile is used to make a consequential decision., No automated result is presented as accreditation, diagnosis, payment approval, or legal advice.
- Reasoning
- The platform uses the declared purpose and approved signals to suggest the next useful action. It does not infer a person’s worth, eligibility, or identity from unrelated behaviour.
- Review
- Required for admissions decisions, credentials, payments, safeguarding, identity, publication, tenant deployment, and other consequential actions.
- Next step
- Review the explanation, choose the supported next action, or use the support route to request human help.
What this does not mean
It does not make an admissions decision automatically, infer sensitive identity, or replace a facilitator. It gives people a useful explanation and routes consequential questions to a human owner.
HAGAN standard: the person affected should be able to understand, question, and safely continue the process.
Inclusive by design
Ready for low-data learning and Braille access
AIOT keeps headings, labels, instructions, progress, errors, and status messages in a semantic reading order. Braille reading mode reduces motion and visual dependency; a screen reader and refreshable Braille display provide the Braille output.
Keyboard firstRegistration, learning, and support paths remain usable without a mouse.
Low-data rhythmClear text, staged pages, and re-entry points help learners continue after interrupted connectivity.
Human supportAccessibility needs and programme questions can be routed to a person rather than hidden behind automation.