AI News
04 Jul 2026
Read 9 min
How AI grading tool for Singapore schools cuts marking time
AI grading tool for Singapore schools cuts marking time to minutes and provides reliable feedback.
An AI grading tool for Singapore schools is turning weeks of marking into minutes, while keeping teachers in charge. Built by three NUS students, the system drafts grades and rich feedback from rubrics and learning goals. Teachers then review and approve. Early pilots show faster, more consistent marking and clearer insights for classes.
When a senior lecturer at the National University of Singapore had to mark essays from nearly 500 freshmen, the task once dragged on for six weeks. With a new AI assistant called Ren, the first pass took about five minutes. More important, he says the feedback is now more consistent and checks ethical frameworks more reliably.
Why teachers needed help
Large classes strain time and consistency
Many teachers work with a team of teaching assistants. Each person may grade a different group of students. Scores and comments can vary. A common starting point helps. Ren gives that baseline, so teaching assistants can focus on nuance rather than starting from zero.Quality feedback drives learning
Students learn faster when comments are clear, specific, and timely. One junior college student used the tool to practise paragraphs for General Paper. She received line-by-line suggestions and fixed a mix-up between human rights and civil liberties. Rapid feedback helped her correct ideas while they were still fresh.How the AI grading tool for Singapore schools works
From submission to first-pass grade
– Students upload handwritten or typed work. – Teachers upload rubrics, syllabus notes, and learning outcomes. – The AI drafts a grade and detailed feedback tied to the rubric.Human-in-the-loop review
– Teachers edit tone, add examples, and fix any missed nuance. – No grade or comment goes to a student until a teacher approves it. – The system learns teacher preferences, from phrasing to mark schemes, so later drafts need fewer edits. By the second or third assignment, most comments are accepted with little change.Actionable reports
After marking, Ren compiles reports for students, assignments, and the whole class. It flags strengths and gaps by topic and question type. Teachers can spot patterns, plan reteaching, and group students who need the same support.Results seen in class
Faster marking, more time for teaching
The speed gain is clear: a short first pass happens in minutes. Teachers then spend their time where it counts—clarifying tricky ideas, extending strong work, and helping students who struggle.Fairer, steadier grading
Because the AI follows the same rubric every time, it reduces swings in scores across classes. Reviewers adjust for context and edge cases, but the baseline stays steady.Deeper, clearer feedback
Students receive concrete suggestions and examples. They can practise in smaller chunks, such as a paragraph at a time, and see quick corrections. That builds better writing and reasoning.Where schools are trying it now
The tool, built by three 23-year-old NUS students—Wong Eu En, Justin Cheah, and Natasha Koh—under the start-up Ren Education, is in pilots across 11 institutions, including NUS and the School of Science and Technology. A full schoolwide roll-out at St Andrew’s Junior College is slated for July. About 40 to 50 educators are involved, and the team expects to support around 5,600 students by mid-year. Schools buy customised annual plans. The start-up also teamed up with Mendaki to support A-level students in literature, history, chemistry, and economics. Hundreds are expected to benefit through tutoring that uses the same feedback engine. Singapore’s education system is also testing other AI tools. On the national Student Learning Space, teachers can use systems like Markly for feedback, Authoring Copilot for lesson ideas and activities, and student-facing feedback assistants. Together, these tools point to a wider shift: faster cycles of work, review, and improvement.Why this approach stands out
AI does draft work; teachers make final calls
Many tools try to replace steps end to end. Ren instead does the heavy lift on the first draft, then hands control back to the teacher. That protects academic judgment and context.Learns each teacher’s style
Because the system adapts to phrasing and grading patterns, it becomes more helpful over time. It reduces rewrite fatigue and keeps feedback aligned with a school’s voice.Works across subjects
– Ethics and humanities: checks argument structure, use of frameworks, and clarity. – Language arts: gives line edits, examples, and tone guidance. – Sciences: evaluates reasoning, method, and explanation quality.How schools can get the most value
Set strong rubrics
Clear rubrics make the AI smarter. Define criteria, point weight, and examples of good work. Tight inputs mean better first-pass outputs.Train the teaching team
Brief markers on how to review AI drafts. Agree on tone, examples, and common fixes to speed up approval.Use the data
– Re-teach topics where many students struggle. – Share anonymised model answers and common mistakes. – Track growth across terms to guide instruction.What this means for students and parents
Students get more timely, specific feedback and can practise in shorter cycles. Parents may see clearer reports that show strengths and gaps by topic, not just a single grade. The system also helps ensure fairness when many people mark the same course. As schools look to cut workload without hurting quality, an AI grading tool for Singapore schools offers a practical path. It speeds up routine marking, raises consistency, and gives clearer insight, all while keeping teachers in control. With careful use and strong rubrics, it can make feedback faster, fairer, and more effective.For more news: Click Here
FAQ
Q: What is Ren and how does it work?
A: Ren is an AI grading tool for Singapore schools developed by three NUS students that drafts grades and detailed feedback from teacher-uploaded rubrics and learning goals. Teachers review, edit and approve every grade and comment before they are released to students.
Q: How much time can Ren save compared with manual marking?
A: In one example, marking essays from 490 students that once took six weeks had a first pass reduced to five minutes with Ren. The AI produces a draft grade and feedback so teachers can finalise assessments much faster.
Q: How does Ren keep teachers in control of grading?
A: Ren follows a human-in-the-loop approach: the AI generates the first-pass grade and comments, but teachers make the final decisions and edits. The system learns teachers’ phrasing and marking preferences over time, so later drafts typically need fewer changes.
Q: What types of student work can the tool handle?
A: Students can submit handwritten or typed assignments on the platform. The AI uses uploaded rubrics, syllabus materials and learning outcomes to generate feedback tied to assessment criteria.
Q: Where is Ren being piloted and how many students or educators are involved?
A: Ren is being piloted in 11 institutions in Singapore, including NUS and the School of Science and Technology, with a full schoolwide roll-out planned at St Andrew’s Junior College in July. The team is working with 40 to 50 educators and expects to support about 5,600 students by July. The start-up also partnered Mendaki to run tutoring sessions that are expected to benefit about 600 A-level students.
Q: How does Ren improve consistency and the quality of feedback?
A: By applying the same rubric and systematically checking for required frameworks, Ren creates a common baseline that reduces variation in scores across markers. It also provides detailed, line-by-line suggestions and class reports that highlight strengths and gaps so teachers can target reteaching.
Q: Which subjects and tasks can Ren support?
A: Ren works across ethics and humanities to check argument structure and framework use, language arts to offer line edits and tone guidance, and sciences to assess reasoning and method. It also supports shorter practice cycles such as paragraph drafting and produces reports to group students for targeted support.
Q: How can schools get the best results from using the AI grading tool for Singapore schools?
A: Schools should set clear rubrics, train teaching teams on how to review AI drafts, and agree on tone and common fixes to speed approvals. They should also use the tool’s reports and data to re-teach topics, share anonymised model answers, and track student growth while subscribing through customised annual plans.
Contents
