In Singapore's competitive higher education landscape, where institutions like the National University of Singapore (NUS) continuously seek innovative ways to enhance teaching and learning, a group of enterprising students has introduced a practical solution to one of educators' most time-consuming tasks. Three NUS undergraduates and recent graduates have developed Ren, an artificial intelligence-powered tool designed to streamline assignment marking while delivering consistent, high-quality feedback.
Addressing Persistent Challenges in Singapore's Education System
Marking student work has long posed significant demands on teachers and lecturers across Singapore's schools and universities. With large class sizes common in modules at NUS and other institutions, educators often spend weeks reviewing assignments, leading to delays in feedback that can hinder student progress. Variations in marking standards among teaching assistants further complicate consistency, particularly in subjects requiring nuanced judgment such as ethics, literature, or essay-based assessments.
The Ministry of Education (MOE) has introduced several AI-supported features on its Student Learning Space platform, including tools for feedback generation. However, many educators still seek solutions tailored specifically to their rubrics and teaching styles. Ren emerges as a complementary innovation born directly from the NUS community, addressing these gaps through a human-in-the-loop approach that keeps teachers firmly in control.
The Origins of Ren Education at NUS
Ren was created by three young innovators connected to NUS: Wong Eu En, a second-year computer science student; Justin Cheah, a fourth-year student in computer science and business; and Natasha Koh, who recently graduated with a degree in information systems. At just 23 years old, the trio founded the edtech startup Ren Education after identifying the marking burden during their own academic experiences and interactions with educators.
Their motivation stemmed from observing how time-intensive grading limited opportunities for personalized student support. By leveraging their technical expertise and understanding of Singapore's rigorous educational environment, they built a tool that processes both handwritten and typed submissions, drawing on uploaded marking rubrics, syllabus materials, and learning outcomes to produce initial drafts.
How the Ren AI Tool Operates in Practice
The process begins when teachers upload their specific criteria and reference materials. Ren then analyzes student submissions and generates draft grades alongside detailed, constructive feedback comments. Educators review every recommendation, making edits to tone, content, or accuracy as needed before final approval and release to students.
A key strength lies in the system's adaptive learning. Over successive assignments, Ren refines its outputs based on individual teacher preferences, such as preferred phrasing or emphasis on certain concepts. This personalization results in teachers accepting 80 to 90 percent of the AI-generated comments without changes by the second or third use.
Beyond individual grading, the platform produces comprehensive reports highlighting class-wide trends, common strengths, and areas for improvement by topic or question type. This data-driven insight supports targeted instructional adjustments.
Real-World Pilots and Measurable Impact
Early deployments demonstrate substantial efficiency gains. In one NUS computing module taught by senior lecturer Lee Boon Kee, marking 490 freshmen assignments previously required six weeks. With Ren, the initial pass now takes approximately five minutes, allowing more time for in-depth review of complex responses and focus on students requiring additional support.
The tool has expanded to pilots across 11 Singapore institutions, including NUS and the School of Science and Technology. It serves around 40 to 50 educators and is on track to reach 5,600 students by July 2026. A full school-wide rollout is scheduled for St Andrew’s Junior College in July.
University-scale testing includes grading 426 individual reports for NUS Computing's IS1108 module, with feedback aligned to actual module teaching practices. These pilots underscore Ren's scalability within higher education settings.
Photo by Pang Yuhao on Unsplash
Partnerships Expanding Reach in Singapore
Ren Education has formed strategic collaborations to broaden access. A notable partnership with Yayasan MENDAKI provides free academic support for A-level students in subjects including literature, history, chemistry, and economics. Approximately 600 students are expected to benefit from enhanced feedback during tutoring sessions.
Schools subscribe via customized annual plans, ensuring the tool integrates smoothly into existing workflows. This model aligns with Singapore's emphasis on practical, scalable edtech solutions that complement rather than replace human expertise.
Student Perspectives on Enhanced Learning
Feedback from users highlights the tool's value for learners. Junior college student Chen Ziling, who used Ren during General Paper tuition, noted that rapid, line-by-line suggestions enabled more effective practice. The platform flagged conceptual misunderstandings, such as distinctions between human rights and civil liberties, providing examples that helped refine her writing in ways limited teacher time sometimes cannot.
Such personalized, timely input supports deeper engagement, particularly valuable in Singapore's fast-paced academic environment where students juggle multiple demands.
Balancing AI Assistance with Human Oversight
Central to Ren's design is the principle that artificial intelligence augments rather than supplants educator judgment. The founders emphasize avoiding fully automated systems, instead positioning the tool as a collaborative aid. Teachers retain final authority, editing outputs to capture nuances the AI might overlook.
This approach mitigates risks associated with AI in assessment, such as potential biases or missed contextual elements, while capitalizing on consistency and speed. It reflects broader discussions in Singapore's higher education sector about responsible AI integration.
Implications for Faculty Workload and Student Outcomes
For academics and administrators at NUS and similar institutions, tools like Ren offer pathways to reallocate time from routine marking toward mentorship, curriculum development, and research. Reduced grading loads may also improve retention and satisfaction among teaching staff.
Students benefit from faster, more consistent feedback that supports iterative improvement. Class reports enable data-informed teaching adjustments, potentially raising overall performance in modules across disciplines.
Future Outlook for AI in Singapore Higher Education
As Ren Education prepares for wider adoption, its trajectory points to continued evolution. Ongoing refinements based on user feedback and expanding institutional partnerships position the startup as a contributor to Singapore's smart nation initiatives in education.
The tool's success illustrates how student-led innovation at universities like NUS can address real-world challenges, fostering an ecosystem where emerging technologies enhance traditional strengths in rigorous, high-quality instruction.
Broader trends suggest increasing integration of such aids, provided they maintain strong human oversight and align with local pedagogical values.
Actionable Insights for Educators and Institutions
University administrators considering similar tools should prioritize solutions that allow customization to specific rubrics and encourage gradual adoption through pilot programs. Training sessions on effective review processes can maximize benefits while preserving academic standards.
Faculty members may explore how AI-generated drafts free capacity for deeper student interactions, ultimately strengthening learning communities within Singapore's higher education institutions.