Report Hub: an AI report builder for early‑childhood school operations.
Paper SOP forms replaced by a full digital report lifecycle — template studio, guided filling, approval routing and analytics — with AI assisting the author, the filler and the approver.
- Field types
- 30
- AI subsystems
- 4
- Approval modes
- 3
- Languages
- 2
Observation checklists, incident reports and training reflections lived in printed SOP documents across the Iyad Perdaus early‑childhood centres. Staff filled them inconsistently, approvals moved by email, and leadership had no live view of what was submitted, pending or overdue across centres.
Report Hub is the centralised reporting engine inside iSMS, the school‑management platform those centres run daily. It carries a report through its whole lifecycle in one place: a drag‑and‑drop template studio with 30 field types — from signatures and file uploads to rating scales and tables — with per‑field properties, branching logic, bulk import of existing forms from CSV or Excel, and versioned publishing. Admins assemble each report in four guided steps (Form Builder, Report Builder, Access Settings, Workflow Settings) with progress saved at any point.
Filling is built for busy staff: focus mode presents one question at a time with live progress, debounced autosave, offline sync for unreliable connections, identity auto‑fill, signatures and rich text — with version history on every submission. Approvals then route single, sequential or parallel, with per‑field remarks, and finish as designed PDF exports.
AI assists at every stage rather than sitting in a sidebar. AI Buddy is a retrieval‑augmented chat assistant grounded in the school’s own SOP library and Knowledge Hub — hybrid retrieval combining dense embeddings and BM25 with reranking, answers streamed in real time with cited sources, voice input and spoken replies in English or Malay, and an “explain this field” helper while a report is being filled. Questions it cannot answer escalate to admins and feed back into the Knowledge Hub.
While staff write, AI writing suggestions review each answer across four independent dimensions — grammar, professional tone, relevance, and missing details — each one tap to apply, edit or dismiss, and tuned to local context from ECDA and MOE acronyms to Malay‑language answers. When an approver flags a field, AI remark resolution drafts a compliant replacement answer that addresses the remark directly. Alongside it, live analytics track submissions per centre and drill into status breakdowns, so leadership finally sees the whole picture as it happens.