The Dubai Schools Inspection Bureau evaluates 200+ private schools across six performance standards. GradeLab feeds directly into the assessment quality, student feedback, and data-for-improvement indicators across three of the six standards. When an inspector visits, your exam data tells the story | not an anecdote, not a spreadsheet, but a live system producing evidence every term.
See Standard MappingEach capability mapped to GradeLab's automated pipeline and the manual legacy process - green means covered, red means a gap.
| Capability | GradeLab (Automated) | Manual / Legacy | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Marks Capture | AI Grading | Feedback Engine | Trends & Reports | Audit Trail | Spreadsheets | Faculty Entry | Cohort Sheets | Manual Recompute | Static PDFs | |
| Assessment Consistency | Rubric-locked | Consistency check | Variance report | Class trends | Override logged | No rubric | Ad hoc | No check | Inconsistent | Opaque |
| Feedback to Students | Per-question | AI feedback | With results | Delivery tracked | Logged | Delayed | Generic | Weeks later | No tracking | No log |
| Data-Driven Teaching | Digital + handwritten | Per-question | Mistake clusters | Topic trends | Item analysis | Manual pivot | Hand-typed | Anecdotal | No trend | Static list |
| Attainment Reliability | Per-student | Rubric-anchored | Bias-reduced | Term trends | Reproducible | Teacher variance | Hand-graded | Bias risk | No trend | Inconsistent |
| Self-Evaluation Data | Live dashboards | Per-student data | Attainment trends | Multi-term | Inspector-ready | Manual pivot | Hand-typed | Inconsistent | Single term | Static list |
| Safeguarding & Compliance | No training | Teacher-in-loop | RBAC + encryption | Audit trail | Dossier ready | Unknown vendor | No oversight | No RBAC | No trail | Retroactive |
| Progress Tracking | Per-student | Per-question | Term comparison | Year trends | Reproducible | Manual pivot | Hand-typed | Inconsistent | No trend | Static list |
| Curriculum Adaptation | Score distribution | AI clustering | Topic weak spots | Per-term | Validated | Manual count | Inconsistent | No standard | Not tracked | Static chart |
GradeLab is not a whole-school solution | it doesn\u2019t teach Arabic, manage behavior, or design curriculum. But on assessment quality, feedback, and data-driven leadership, it produces the system-generated evidence that inspectors evaluate.
The core domain | inspectors assess teaching quality, assessment practices, use of data, and feedback to students
| What Inspectors Evaluate | GradeLab Feature | Evidence Produced |
|---|---|---|
| Quality of assessment | consistency, standardisation, alignment to curriculum | Rubric EnforcementConsistency CheckAI Confidence Scoring | Every script graded against the same rubric; consistency-check reports showing grading variance across the class; teacher override audit trail |
| Use of assessment data to inform teaching | Common Mistakes ClusteringLearning InsightsAI Q&AItem Analysis | AI-clustered recurring mistake patterns; "Which topics are students consistently weak on?" | answered from live exam data; per-question discrimination index |
| Quality of feedback to students | Per-Question FeedbackOverall FeedbackStructured Feedback | Strengths, mistakes with fixes and reasons, improvement pointers | per question, per student; feedback delivered with results, not weeks later |
| Curriculum adaptation from assessment results | AnalyticsPer-Question AnalysisGrade Distribution | Score distributions, question difficulty classification, topic-level weak spots | teachers adjust instruction based on actual performance data |
Attainment and progress in Islamic Education, Arabic, English, Mathematics, Science, and other subjects
| What Inspectors Evaluate | GradeLab Feature | Evidence Produced |
|---|---|---|
| Reliability of attainment measurement | Consistency CheckAudit TrailStudent Performance | GradeLab doesn’t teach the subjects, but the accuracy of internal assessment data | rubric-enforced, bias-reduced | means attainment measurement is reliable, not teacher-by-teacher guesswork |
| Progress tracking across terms | Exam ReportStudent HistoricalClass Statistics | Term-by-term performance data; per-student progress visible across exam cycles; British-curriculum schools can feed predicted grades from live internal data |
Governance, self-evaluation, improvement planning, partnerships, staffing, and resources
| What Inspectors Evaluate | GradeLab Feature | Evidence Produced |
|---|---|---|
| Use of data for self-evaluation and improvement | Longitudinal DashboardsLearning InsightsAI Q&A | "How have our Year 11 science results trended across three terms?" | answered from live data; inspector sees evidence-based leadership, not anecdotal governance |
| Management of risks (including data protection) | RBACAudit LogsData EncryptionTeacher-in-Loop | No model training on student data; full audit trail; RBAC with granular permissions; encryption at rest and in transit; teacher owns every final mark |
Safeguarding, health and safety, SEND support, and online safety
| What Inspectors Evaluate | GradeLab Feature | Evidence Produced |
|---|---|---|
| Safeguarding and online safety | Data ComplianceTeacher-in-LoopRBAC | Student work never trains models; human owns every final mark; audit trail per script; the compliance dossier (DPA, DPIA, sub-processor register) feeds the school’s safeguarding evidence file |
Moral education, social responsibility, curriculum adaptation, and enrichment
| What Inspectors Evaluate | GradeLab Feature | Evidence Produced |
|---|---|---|
| Personal and social development | N/A | GradeLab doesn’t teach personal development outcomes or design curriculum. These standards are outside our scope. |
Inspection evidence that took weeks of manual compilation per visit now happens automatically - zero spreadsheets, zero retrospective data gathering
3/6 Standards
DSIB standards supported with system-generated evidence
Per-script audit trails with override reasoning prove assessment evidence is system-generated - inspector-verifiable on demand
100%
Inspector-ready
Per-script audit trail
Rubric enforcement records
Feedback delivery logs
Compliance dossier
Longitudinal dashboards across terms, feedback delivery tracking, and attainment trends - all visible, all inspection-ready
Term 1 → Term 3
Year 11 science attainment trend
For a Dubai school, the real AI risk isn\u2019t using the technology | it\u2019s using a vendor who can\u2019t answer the DPO\u2019s questions. GradeLab arrives with the dossier your data protection lead will ask for.
Contract-grade commitment: student work is used solely for that school’s grading. Never trains models. Put in the agreement.
Every final mark is owned by a human. AI proposes; teachers review, override, and approve. The audit trail records every action.
Per-question scores, AI reasoning, confidence levels, overrides with reasons | every grading action timestamped and attributable.
Data encrypted in transit and at rest. RBAC with 60+ granular permissions. SSO/MFA. Dedicated encryption for enterprise deployments.
British-curriculum schools benchmark against UK GDPR, ICO Children’s Code, and KCSIE standards. GradeLab is built to the highest overlapping standard.
Signature-ready DPA, security whitepaper, architecture diagram, sub-processor register, pen-test summary, and pre-filled DPIA template | before the first meeting.
When the DSIB inspector asks about safeguarding in digital assessment, your leadership team can demonstrate that you selected an AI vendor with a proactive compliance posture | not one that scrambled to answer questions retroactively. The dossier is part of your school\u2019s evidence file for Standard 5.
UAE-region hosting, SOC 2, and ISO 27001 status are stated truthfully with current status plus dated roadmap | never exaggerated. One discovered exaggeration permanently kills a trust sale with a KHDA-governed school. UAE data compliance specifics are verified against PDPL 45/2021, Wadeema 29/2016, and current executive regulation status before appearing in any written commitment.
GradeLab feeds directly into three of the six DSIB performance standards: Teaching and Assessment (Standard 3), Leadership and Management (Standard 6), and Protection/Care (Standard 5). It produces system-generated evidence | rubric consistency reports, feedback delivery logs, longitudinal dashboards, and audit trails | that inspectors evaluate. When the inspector asks about assessment quality, you show a live dashboard, not a spreadsheet.
GradeLab arrives with a proactive compliance dossier: signature-ready DPA, security whitepaper, architecture diagram, sub-processor register, pen-test summary, and pre-filled DPIA template. No model training on student data is a contract-grade commitment. Full audit trail per script, RBAC with 60+ permissions, encryption at rest and in transit, and teacher-in-the-loop on every final mark.
Yes. GradeLab benchmarks against UK GDPR, ICO Children’s Code, and KCSIE standards | the highest overlapping standard. British-curriculum schools can feed predicted grades from live internal data, track term-by-term progress, and produce attainment evidence aligned with both DSIB and British curriculum requirements.
GradeLab does not cover Personal Development (Standard 2) or Curriculum Design (Standard 4) | moral education, social responsibility, curriculum adaptation, and enrichment. We are transparent about where we contribute and where we don’t. The same honest-fit discipline we bring to every framework.
GradeLab is designed to align with PDPL 45/2021 and Wadeema 29/2016. UAE-region hosting, SOC 2, and ISO 27001 status are stated truthfully with current status plus dated roadmap | never exaggerated. UAE data compliance specifics are verified against current executive regulation status before appearing in any written commitment.
Walk into your next DSIB inspection with system-generated evidence - rubric consistency, feedback delivery logs, and longitudinal dashboards - not a last-minute spreadsheet.
Save 100+ hours per month
AI-powered accuracy & consistency
Instant student feedback
Easy LMS integration
