Drafts and revisions
Compare saved versions and replay the recorded development of the assignment.
Student writing · Recorded AI help · Teacher reviewBuilt first for UAE schools
Turn students’ writing and AI use
into evidence teachers can review.
Reasonleaf is a prototype assignment workspace. Students write and save drafts; it records AI help used inside the workspace. Teachers review that history, ask students to explain their choices and keep feedback with the submission.
Book a 20-minute discovery call See the prototype review record 20-minute educator conversation · No purchase commitment
01 / Teacher
Configure the assignment and permitted AI help.
02 / Reasonleaf
Record drafts, sources and in-workspace AI conversations.
03 / Teacher
Review the evidence and document feedback.
The output / An assignment review record01 02 03 04
Reasonleaf brings the writing history into the teacher’s review.
In the prototype, the final submission sits beside saved drafts, recorded assistance and the student’s explanation. Teachers can compare versions, choose a follow-up question and save feedback with the work.
Recorded AI conversations
Read the requests and responses captured inside the assignment workspace.
Sources and paste events
Inspect source entries and recorded pastes. Ask for context where the origin is unknown.
Student explanations
Read why a student changed their argument and what assistance they declare.
Reasonleaf records activity inside its assignment workspace. It does not monitor other apps or prove independent authorship.
Teacher reviewWorking prototype
After reviewing the recordThe teacher records the decision.
The teacher records the decision.
The student can explain the work.
A draft written elsewhere, language support or a pasted passage needs context. Reasonleaf includes the student’s explanation and the teacher’s feedback in the record. No automated misconduct verdict is produced.
From submitted work to teacher feedback
Review the submission and the work behind it.
See saved drafts, recorded AI assistance and the student’s explanation together. Use that record to ask a focused follow-up and document feedback.
What we want to test with your teachers: whether that record makes gathering evidence and completing a review easier.
Reasonleaf currently uses a separate prototype workspace. Connections to existing school platforms remain planned.
Book a discovery callFor UAE principals and educators
How do your teachers review AI-assisted work?
Tell us what happens at your school. We’re speaking with educators to understand where reviewing student work takes too long, or leaves questions unanswered.
After the conversation
We’ll send a short summary of your review workflow, the gaps we identify together and whether a focused prototype could help.
No preparation, pupil records or purchase commitment needed.
School discovery callOne assignment.
One assignment.
An honest conversation.
20 minutesOnline
How your teachers review work
Talk through a recent assignment and the tools they use today.
Where AI makes that harder
Discuss missing evidence, unclear rules or extra review time.
Whether Reasonleaf could help
Agree if there is a specific problem worth exploring together.
Opens Google Calendar in a new tab.
Choose an available time in your timezone.
English, maths, science and beyond
What would teachers need to review in your subject?
Explore the proposed workflows: what students would submit, what teachers would review and what still needs building. The current prototype starts with writing.
Example assignment / English
Write an argument about water conservation.
Review how an argument developed alongside the final submission.
Writing prototypeRich annotation and quotation-to-source links remain planned.
- Students do
- Plan the argument, write drafts, cite sources and explain revisions.
- Teachers permit
- Choose whether brainstorming, vocabulary explanations or draft feedback are permitted.
- Reasonleaf records
- Draft checkpoints, recorded AI conversations, source entries and the student’s reflection.
- Teachers review
- Compare the argument across drafts and discuss a specific revision with the student.
Subject examples describe the intended workflows. The current prototype demonstrates writing, draft checkpoints, recorded assistance and teacher review. Specialist subject tools are labelled as planned.
Handwritten maths / proposed workflowStudents solve on paper.
Students solve on paper.
Teachers review the working.
Keep paper in the classroom. The proposed workflow connects a submitted page to a teacher’s annotation and, where needed, an observed check of understanding.
Fictional working / typed transcription
Solve the two equations
2x + y = 11First equationx − y = 1Second equation3x = 12Teacher selects this stepx = 4, y = 3Submitted answer
Teacher annotation
Illustration of a planned review workflow, not an uploaded student page or a working maths feature.“Why does adding the equations eliminate y?”
- 01
Students upload their working.
A photo or scan is attached to the question, with any outside help declared. It is labelled submitted working, not verified independent work.
- 02
Teachers annotate a step.
The original image stays available. A teacher can request an explanation of a particular calculation or change of method.
- 03
Teachers check understanding.
If needed, the student explains the step or solves a variation under observation. A typed explanation completed at home can also be AI-assisted.
- 04
Teachers record their observation.
Keep the explanation, prompting required and next teaching step with the assignment. A difficult follow-up does not automatically establish misconduct.
Outside the assignment workspace
What if an AI assistant is open on another tab or laptop?
Reasonleaf cannot see another device or an unrelated AI chat. A student can copy an external answer by hand or type it into the workspace. Neither handwriting nor typing history establishes where that reasoning originated.
Recorded activity
Shows what happened inside Reasonleaf. An upload timestamp shows when a page was submitted.
Student declaration
Describes the help a student says they used. It is not an independently verified activity record.
Teacher observation
Documents understanding demonstrated in that setting. It does not prove authorship of the original homework.
How the evidence is collected
Interactive illustration · Fictional student workReasonleaf records the draft, the AI help and the revision.
A more water-wise school.
How could our school use water more responsibly?
Schools in Dubai should do more to save water. A weekly water audit would help students see where water is being wasted and decide which changes would make the biggest difference.
A general claim becomes a specific, testable proposal.
Draft checkpoint09:31
03 / 04
Compare what changed.
Choose a stage above to exploreBefore students start writingTeachers choose the AI assistance
Teachers choose the AI assistance
inside each Reasonleaf assignment.
Reasonleaf shows the selected rule in the student’s workspace and keeps that policy version with the submitted work.
Example assignment setting / Guided AI
Ask for explanations and feedback. Develop your own argument.
What students can do
Students can explore ideas, request feedback and plan research within the assignment’s selected mode.
What Reasonleaf keeps for review
Teachers can inspect the recorded guidance, later revisions and the student’s explanation.
Illustration of the prototype’s six modes. These are school-defined settings, not certified curriculum policy packs. Live AI behaviour still requires evaluation.
From an assignment to a school-wide system
Connect school policy to the assignments teachers create.
Assignment settings work in the prototype. The next layer lets schools maintain reviewed policy packs, define exceptions and control who can access the records.
Carry the applicable rule into the assignment.
Planned policy packs connect school, curriculum, awarding body, subject and assessment requirements. Teachers would start from a rule approved for the subject, year group and assessment type, and request exceptions where needed. Practice and assessment can permit different help.
Give each person access to the right records.
Planned school identity and role permissions separate student, teacher and leadership access. These controls are required before real pupil use.
Retain the policy, evidence and review decision.
The prototype exports an assignment record. Production adds reliable shared storage, audit history, school-controlled retention and deletion.
What the school should measure
Compare total assignment setup, review and follow-up effort with current practice. Check whether students understand the permitted help and whether teachers can document useful decisions. Time savings have not yet been established.
Beyond the first assignmentExplore the full five-module product vision
The complete visionRecord student work.
Record student work.
Apply policy. Support teacher review.
Five connected modules explain what Reasonleaf does now and what the school platform must add next.
Write
Keep drafts and revisions with the submission.
Students write, save checkpoints, add sources and explain changes. Teachers can compare the development of a piece with its submitted version.
Working prototype
Start with an assignment
AI
Set the AI help allowed for each task.
Choose an assignment’s AI mode and record assistance used inside the workspace. The demo uses labelled rehearsal responses; live OpenAI requires configuration and evaluation.
Working prototype
OpenAI integration
Integrity
Review evidence and document the decision.
Inspect drafts, recorded assistance, sources and the student’s explanation together. Add feedback and export the review record. Similarity evidence remains planned.
Core review prototype
Human judgment at the centre
Policy
Apply curriculum and school rules to assignments.
Planned policy packs connect school, curriculum and assessment requirements, with version history and approved exceptions. Assignment-level AI modes work in the prototype.
Planned capability
School-wide consistency
Trust
Control access, retention and school oversight.
Planned school identity, role permissions, retention, deletion and audit controls make the records suitable for institutional use. These are requirements before pupil data.
Planned capability
Institutional readiness
The first workflow connects AI rules, student work and teacher review. Curriculum policy, similarity and institutional controls extend it.
Where we are today UAE-first product prioritiesBuild policy packs for UAE curricula.
Build policy packs for UAE curricula.
Evaluate writing and AI support in Arabic and English.
A British coursework task, an IB essay and a CBSE project can require different assistance rules. We plan to encode reviewed requirements into assignment settings, with Arabic and English workflows evaluated alongside the teachers who use them.
Assignment rules and review evidenceقواعد الواجبات وأدلة المراجعة
Planned coverage includes British, IB, CBSE, CISCE, American and UAE MoE pathways. Arabic writing direction is demonstrated. Full localization, reviewed policy packs and bilingual AI evaluation remain planned.
Why this matters in the UAEPublic initiatives informing our approach
AI literacy, in everyday subjects.
The KHDA, DP World Foundation and MIT RAISE programme brings responsible AI learning into core subjects, including English and Arabic.
Dubai Media Office — AI literacy, in everyday subjects.February 2026Independent thinking stays central.
ADEK’s AI literacy initiative connects classroom practice with safeguards for privacy, cognitive agency, academic integrity and human oversight.
ADEK — Independent thinking stays central.September 2026A national commitment to AI learning.
The Ministry of Education introduced AI into the public-school curriculum from KG to Grade 12 for the 2025–26 academic year.
Emirates News Agency — A national commitment to AI learning.May 2025These initiatives provide context for our work. Reasonleaf is an independent project and is not endorsed by these organisations.
Before we talkYour questions,