Education innovation often starts with a simple idea: that better understanding how students learn can transform how we teach. For Nirmal Patel, that idea led to the creation of Smart Paper, an award winning tool that brings the power of AI to handwritten student work. Developed to capture and analyze learning at scale, Smart Paper makes it possible to see patterns in how students think, make mistakes, and grow. From state assessments to everyday classrooms, Patel’s work is helping educators unlock new insights from data that has long lived on paper. In this 5 Questions interview, he shares how Smart Paper began, what it reveals about student learning, and his vision for the future of AI in education.
What is the nature of your work?
I led the development of Smart Paper, a tool designed to digitize handwritten student work at scale and analyze learning patterns from that data. The project began with the idea that valuable insights into how students think, especially in math, are hidden in handwritten responses. Smart Paper allows teachers to capture students’ written work and uses AI to recognize handwriting, grade responses, and identify misconceptions. We have implemented it in large-scale settings for multiple-choice questions, such as state assessments in India, where millions of students participate. With new advances in AI, we are now beginning to digitize handwriting. The technology supports multiple languages and subjects, from math and English to social science, and helps generate detailed learning analytics that inform teacher training, student feedback, and curriculum design.
Why is this work important?
Most educational data today comes from digital platforms, but the richest learning data found within students’ handwritten work has been largely invisible. By digitizing and analyzing this paper-based work, Smart Paper gives educators a clearer picture of how students think, where they struggle, and how to help them. The tool enables a broader understanding of misconceptions and provides immediate feedback for teachers. It also opens new pathways for learning science research by analyzing students’ handwritten work to reveal patterns in how they reason and make systematic errors. Beyond improving instruction, this work helps bridge the gap between technology and traditional classrooms, especially in regions with limited access to digital tools.
What’s been the biggest surprise so far?
One surprise has been how well AI can now interpret even the messiest handwriting – sometimes better than humans. Seeing the system accurately read extremely unclear responses and grade them correctly has been remarkable. Another unexpected insight came from the data itself: when analyzing thousands of math and writing samples, we saw that students across grades make surprisingly similar mistakes. That discovery reinforces the idea that large-scale handwritten data can reveal deep insights about human learning and cognition. Finally, we have learned that technology alone isn’t enough; successful implementation depends heavily on people. Training teachers to capture high-quality scans is as critical as improving the algorithms that read them.
Where do you see your work in five years?
In five years, I envision Smart Paper being used globally across subjects, languages, and education systems to create a shared “ImageNet for education.” The goal is to build a large, open dataset of handwritten student work that supports better AI tools, more personalized instruction, and deeper research into how people learn. We are already expanding handwriting recognition across languages like Hindi and English and plan to extend to others. Future development will also focus on connecting student assessments to digital feedback, parent-teacher reports, and targeted lesson plans. Ultimately, I hope the platform will help researchers and educators worldwide understand learning more precisely and design AI that learns math and language like humans do.
What else should people know?
Smart Paper was made possible through support from the Tools Competition, which provided critical funding when early partnerships with government schools faced delays. That support allowed us to focus on innovation rather than costs and helped demonstrate what’s possible when research and implementation align. Today, Smart Paper is being used in thousands of schools and by millions of students. It’s deeply rewarding to see how technology built for social impact can empower educators, scale rapidly, and reveal new dimensions of learning. Every time I visit a school and recognize students whose work our system has analyzed, it reminds me why this work matters.
