JC SEN · JC SEN Portfolio
A calm, private growth portfolio for neurodivergent primary pupils — with a small, human-approved AI helper.
Draft content 8 item(s) marked "to be confirmed" by the programme office.
About the Platform
JC SEN Portfolio is the in-app name of the JC SEN AI-Powered Student Portfolio Platform. It is one part of a five-year programme initiated by The Hong Kong Jockey Club Charities Trust: *Fostering an Inclusive and Empowering Learning Environment for Neurodivergent Primary School Students*. This page is not their official site.
The Platform does three things: a structured profile for each student, growth tracking across many areas, and data entry on a phone, anywhere. It is strength-based and never diagnostic. The AI is only a small helper — every core workflow works without it.
Built by the delivery team — organisation name to be confirmed.
Demonstration · synthetic data only — no real children.
Seven role-scoped views
SEN support is everyone's job — this is the Whole School Approach. That is why the Platform has seven role-scoped views. Each person sees the same evidence, in the way that is right for them.
Students always use the Platform with a supervising adult. School Admin manages metadata only and cannot open any child's portfolio. NGO Coordinator and Researcher see de-identified figures only — never names, never media.
How data is protected
The Personal Data (Privacy) Ordinance (Cap. 486) is the legal minimum the Platform must meet. All portfolio content is treated as confidential personal data of students, teachers and staff. The Platform collects only what a strength-based, non-diagnostic portfolio needs — there is no diagnosis, label, eligibility or risk-score field. Each school's data is kept separate; access is denied by default and granted by role.
The consent flow in this demo — a one-time invite link, with full / partial / pending / declined levels — is the delivery team's design. The final consent model and notice wording are to be confirmed with the programme office before any real data is collected.
- Consent first — Nothing reaches a family until a parent gives consent.
- Partial consent — Under partial consent, video stays with the school. Withdrawing consent empties the parent view at once, with a plain explanation.
- Small cohorts hidden — Small SEN cohorts are suppressed (k-anonymity) for NGO & researcher views. Suppression is shown as 'cohort < k', never silent.
- Audit trail — Every AI call, consent change and share is recorded in the audit trail. Entries carry student ids, not names.
- One school, one space — A school never sees another school's records. Cross-school and cross-role attempts are blocked and logged.
- Child-safe by design — The child's view is read-only and strengths-only: no inputs, no AI, no numbers that read like grades, no labels, no alerts, and no path to sensitive data.
- Demonstration data — Every figure in this demo is computed live from a synthetic demo database. No real child, parent, teacher or school is present.
The AI stance
AI is a helper on top of trusted portfolio records — it is never the product. Every core workflow works with the AI switched off. Its wording is strength-based and respectful, and follows an approved bilingual terminology guide. When a teacher accepts a suggestion, it becomes a teacher-approved record with the label AI-assisted · teacher approved. A rejected suggestion is discarded. Nothing is saved until the teacher accepts.
Video 'insight' signals in this demo are provisional (model: spec pending). They are a prompt for the teacher's judgement, never a diagnosis — and they are not validated.
- Suggests, never decides — It suggests — it never decides.
- Never diagnoses — It never diagnoses, labels, or scores a child.
- Teacher approves first — Nothing it writes reaches a student or parent until a teacher approves it.
- Everything logged — Every AI action is logged. Each output records the model, prompt and version used.
- Works with AI off — The whole Platform works with the AI off. A per-feature switch removes every AI feature while the portfolio flow continues.
- Grounded in evidence — Drafts come only from the child's own records and the approved framework. Every summary links the records it used; missing data is shown, not filled in.
The team
Names and roles will be listed once the programme office confirms them. No individual is named on this demonstration page.
*Roster to be confirmed.*
| To be confirmed | Programme Lead, programme office — to be confirmed |
|---|---|
| To be confirmed | Delivery team lead — to be confirmed |
| To be confirmed | Research partner liaison — to be confirmed |
| To be confirmed | NGO partner liaison — to be confirmed |
Partners
The Platform is designed for a multi-year, multi-school programme, with each school in its own separate space. Partner names and logos will appear here only with written approval from the programme office. Until then, this section is a placeholder — to be confirmed.
Contact
For questions about the programme or this demonstration, please contact the programme office. Contact details are to be confirmed.
Demonstration · synthetic data only — no real children.
👥 Seven role-scoped views
- Teacher / SENCO — A class board, evidence capture, and AI-drafted reports.
- Student — A child-safe view of their own work — no scores, no labels.
- Parent / Caregiver — A calm summary of what the school has shared about their child.
- School Admin — Manage roster, staff, classes and consent — metadata only.
- Programme Admin — A live programme-wide dashboard, settings and AI governance.
- NGO Coordinator — De-identified cohort roll-ups — no names, no media.
- Researcher — Threshold-suppressed aggregate measures for study.
🛡 How data is protected
- Nothing reaches a family until a parent gives consent.
- Under partial consent, video stays with the school.
- Small SEN cohorts are suppressed (k-anonymity) for NGO & researcher views.
- Every AI call, consent change and share is recorded in the audit trail.
✨ The AI stance
- It suggests — it never decides.
- It never diagnoses, labels, or scores a child.
- Nothing it writes reaches a student or parent until a teacher approves it.
- Every AI action is logged.