Gaurav Jain, CEO & Co-Founder, Funscholar Innovations Private Limited, Kolkata, West Bengal, talks about how AI is changing the way students learn and teachers assess, making education more personalised, efficient and focused on individual progress.
From marks to meaning
For decades, student assessment has followed a familiar pattern. Students take up examinations, teachers spend countless hours marking answer scripts, and marks are recorded as the primary measure of academic performance. But this system has often failed to answer a far more important question: What has the student actually learnt?
Traditional assessment methods provide limited insights into a student’s understanding, strengths and weaknesses. Teachers, burdened with administrative responsibilities and heavy workloads, often struggle to offer detailed, individualised feedback. Consequently, students receive marks without understanding where they went wrong or how they can improve.
Assessment has long been regarded as the final stage of learning—a process that ends with marks, rankings and report cards. However, the growing emphasis on personalised education highlights the need for assessment systems that actively contribute to learning rather than simply measuring it.
Recognising these longstanding challenges, the education technology company Funscholar has developed ClassMap, an AI-powered assessment platform designed to transform the way schools evaluate learning. The platform aims to make assessment an ongoing learning process by providing personalised feedback, identifying learning gaps and supporting teachers with intelligent automation.
ClassMap seeks to address this need by combining artificial intelligence with educational insight. Through automated grading, detailed feedback, adaptive learning pathways and data-driven classroom analytics, the platform aims to reduce teachers’ administrative burden while giving students a clearer understanding of their academic progress.
Rather than viewing marks as the ultimate objective, the platform encourages schools to focus on continuous improvement, conceptual understanding and individual learning needs. Ultimately, the future of education may not depend on replacing traditional assessment, but on enriching it with the feedback, insights and adaptability that every learner deserves.
A vision takes shape.
The vision behind ClassMap stems from years of experience in both engineering and educational technology. After completing an engineering degree in Bengaluru, followed by a Master’s degree from IIT Kanpur and an Executive MBA from IIM Kolkata, the founder of Funscholar dedicated his career to making education more experiential and engaging.
For more than twelve years, the company has specialised in robotics and artificial intelligence, helping schools introduce practical, hands-on learning experiences into classrooms across India. Today, Funscholar works with more than 4,000 schools while collaborating closely with central and state government institutions.
Building on this extensive experience, the company has expanded its focus to assessment through ClassMap, a platform that seeks to redefine how schools measure student learning by combining artificial intelligence with personalised educational insights.
Beyond examination scores
Assessment has traditionally focused on examinations and marks. Students write papers, teachers evaluate them, and the final score determines academic success. However, this process often overlooks the most valuable aspect of education—the learning journey itself.
According to the philosophy behind ClassMap, students should not be judged solely by examination results. Instead, assessments should reflect their ability to understand concepts, apply knowledge, and continuously improve.
Beyond the red pen
One of the most significant issues is the sheer workload teachers face every day.
A single teacher may handle multiple classes and hundreds of students, while simultaneously preparing lesson plans, assigning homework, conducting examinations and marking answer scripts. By the end of the school day, there is little time or energy left in her for providing meaningful academic guidance.
Marking itself becomes a repetitive administrative task rather than an opportunity to improve student learning. Teachers are expected to evaluate large numbers of scripts within limited timeframes, making detailed feedback almost impossible. They are often only able to indicate whether an answer is correct or incorrect before moving on to the next paper.
On average, teachers spend between one and two hours every day marking homework, classwork and assignments. During examination periods, the workload increases substantially as hundreds of answer scripts must be evaluated within tight deadlines.
According to the company, ClassMap can save teachers around 10 to 12 hours each week, while reducing marking time by approximately 100 hours during examination seasons. Over the course of a year, the total time saved can reach 300 to 400 hours. These hours can instead be devoted to lesson preparation, classroom interaction, and providing direct academic support to students.
Beyond the marks
Another challenge is the overwhelming emphasis placed on examination performance.
Students frequently associate academic success with marks rather than understanding. The pressure to perform comes not only from schools but also from parents, peers and society. This examination-centred culture often encourages memorisation rather than conceptual learning. Students become focused on achieving high scores instead of understanding why mistakes occur or how they can strengthen their knowledge.
The absence of meaningful feedback means that examinations become an endpoint rather than a stepping stone for further learning.
Feedback fuels growth
The greatest weakness of conventional assessment is the lack of continuous feedback. Consider a mathematics problem answered incorrectly in an examination. In many classrooms, the teacher simply awards zero marks or places a cross beside the incorrect answer. When students review their papers later, they know only that the answer was wrong—they rarely understand why.
To obtain clarification, students must either approach their teacher individually or seek assistance outside school through tuition. Given crowded classrooms and demanding teaching schedules, personalised explanations are often impractical. As a result, many students continue making the same mistakes because they never receive clear guidance on how to improve. When students understand the reason behind an error, they are far more likely to correct it in future assessments.
The limits of OMR
Many schools have adopted digital assessments using Optical Mark Recognition (OMR) sheets to simplify evaluation. While these assessments reduce manual marking, they introduce another limitation. Typically conducted only once or twice during an academic year, OMR-based tests provide a snapshot of student performance at a single point in time. They indicate whether answers are correct or incorrect but offer little insight into the student’s overall learning journey.
More importantly, they fail to identify how understanding evolves throughout the year. Schools may know that a student performed poorly in a particular examination, but they cannot easily determine whether the difficulty stems from conceptual misunderstandings, recurring calculation errors, or specific topics requiring additional attention.
Beyond one-size-fits-all
One of the fundamental principles behind personalised education is that every student learns differently. Some students grasp concepts quickly, while others require additional practice. Individual learners also struggle with different topics.
Despite these differences, classrooms generally assign identical homework, worksheets and revision exercises to every student. This approach ignores individual learning needs and makes it difficult for students to receive targeted support.
Although schools have adopted various educational technologies over the years, many of these systems solve isolated problems without integrating effectively. Teachers also face practical challenges in learning multiple digital platforms while managing already demanding schedules. The result is that technology often becomes underutilised, leaving personalised learning largely unrealised despite its recognised importance.
The ClassMap advantage
Recognising the limitations of conventional assessment systems, Funscholar developed ClassMap with a simple objective: transform assessment from a scoring exercise into a meaningful learning experience. Rather than replacing teachers, the platform is designed to support them by automating routine tasks, generating actionable insights and enabling more personalised instruction.
The platform addresses three key areas simultaneously. It assists teachers in creating assessment content, automates the evaluation of student responses and generates detailed feedback that helps both students and educators understand learning gaps.
Supporting every educator
To create high-quality assessments, teachers should give considerable preparation, design questions that align with curriculum objectives while incorporating different formats, including multiple-choice questions, case studies and competency-based exercises.
This preparation can consume a significant portion of a teacher’s working day, particularly when classes are scheduled back-to-back. ClassMap simplifies this process by generating questions aligned with board requirements.
The platform also automates the evaluation of handwritten assignments, homework, classwork, and examinations. Teachers receive evaluated responses supported by consistent grading criteria and spend the time on lesson planning, classroom interaction, and addressing individual student needs.
Understanding every answer
ClassMap has been developed to evaluate handwritten answers. The system analyses students’ written responses using AI trained on established marking rubrics supplied either by examination boards or individual schools. It applies consistent grading standards while identifying missing points, incomplete explanations and partially correct responses.
This enables the platform to award marks more accurately for subjective answers. For example, if a student solves part of a mathematics problem correctly but makes an error in the final calculation, the system can allocate partial marks while clearly explaining why marks were deducted.
A major strength of ClassMap lies in its ability to evaluate handwritten answers with a high degree of consistency. The platform has been trained using more than 10,000 answer scripts collected from students with varying handwriting styles and from different regions. This extensive training enables the AI to recognise handwriting ranging from exceptionally neat to difficult-to-read scripts.
Currently, the platform supports English and Hindi, with additional regional languages planned for future releases. By accommodating diverse handwriting styles, the system ensures that schools can adopt AI-assisted assessment without requiring students to alter how they write examinations.
Understanding every mistake
One of ClassMap’s defining features is its emphasis on personalised feedback. Traditional assessment ends once marks have been awarded. Students know whether they answered correctly, but they often remain uncertain about the underlying mistake.
ClassMap changes this by providing detailed explanations for every response. Each incorrect or partially correct answer is accompanied by feedback describing what was missing, which concept requires further attention, and how the response could be improved. This transforms assessment into an ongoing learning opportunity rather than a final judgement.
Instead of focusing exclusively on examination scores, students begin to understand the reasoning behind their mistakes. Over time, this encourages deeper conceptual learning rather than rote memorisation. The platform also benefits parents, who gain a clearer understanding of their child’s academic strengths and areas requiring additional support.
The platform is currently being used in 25 schools, following pilot programmes conducted in more than 55 schools. According to the company, schools have observed an average improvement of around 20 per cent in students’ academic performance after introducing personalised AI-generated feedback.
The improvement is attributed not simply to automated grading but to the quality of feedback students receive. Instead of seeing only their marks, learners understand precisely where they made mistakes and how those mistakes can be corrected.
Interestingly, schools have reported that students themselves increasingly request that their answer scripts be evaluated using ClassMap because they value the detailed explanations accompanying their results. Parents have also welcomed the additional transparency.
Every child, unique
No two students learn in the same way. Some require additional support in algebra, while others struggle with geometry, trigonometry, or problem-solving techniques. Recognising these differences, ClassMap introduces personalised adaptive learning.
After evaluating assessments, the platform identifies each student’s individual learning gaps. It then generates customised worksheets and practice questions specifically targeted at those areas. A student requiring additional practice receives a different set of exercises from another student who needs support in a different part.
The system also adapts question difficulty according to each learner’s progress. As students demonstrate improved understanding, subsequent questions gradually become more challenging, creating a continuous pathway for academic growth rather than a fixed level of practice.
Reading classroom patterns
ClassMap provides teachers with valuable classroom-level insights. Teachers receive analytical reports showing which concepts caused the greatest difficulty across an entire class. For example, the system may reveal that a large proportion of students struggled with geometry, while another topic was well understood. Armed with this information, teachers can revisit challenging concepts, modify lesson plans, and allocate additional classroom time where it is most needed. These insights also extend to school leaders, allowing principals and academic coordinators to monitor learning trends and evaluate teaching outcomes.
Reports that reveal
The platform generates comprehensive reports that go well beyond traditional mark sheets. Rather than presenting only total scores, reports identify precise areas requiring improvement. Students can see exactly which concepts caused difficulty, whether errors resulted from calculation mistakes, incomplete reasoning, or misunderstandings of specific topics.
Question-by-question feedback explains both strengths and weaknesses, helping students understand not only what they answered incorrectly but why. For schools following different curricula, including the International Baccalaureate (IB), the platform supports varied marking schemes while maintaining detailed explanatory feedback.
Learning comes alive
ClassMap also encourages a shift away from memorisation by generating contextual, application-based questions. The system creates case-based and data-driven problems linked to real-life situations. A mathematics exercise, for instance, may ask students to calculate dimensions or solve practical scenarios instead of performing isolated calculations.
This approach encourages analytical thinking, problem-solving and conceptual understanding, helping students appreciate how classroom knowledge applies beyond examinations. Such assessments develop critical thinking rather than rewarding memorised answers.
Accuracy meets fairness
One of the most common concerns surrounding AI-assisted assessment is grading accuracy. According to Funscholar, ClassMap currently achieves an accuracy rate of approximately 95 to 96 per cent when compared with human evaluation. The company emphasises that assessment is never entirely objective, even among experienced teachers. Two teachers evaluating the same answer script may award slightly different marks based on their interpretation of the response. Such variations are considered normal within educational assessment.
ClassMap minimises this inconsistency by applying the same marking criteria to every script. Its accuracy depends largely on the quality of the marking rubrics provided by schools or examination boards.
Tailored intelligence
Rather than relying on a single large language model (LLM), ClassMap employs a combination of multiple language models alongside its own specialised Small Language Model (SLM).
The SLM is customised according to the needs of different educational sectors. Separate models can be configured for K–12 schools, coaching centres, higher education institutions and professional courses, ensuring that grading reflects the assessment style, marking schemes and curriculum requirements of each institution.
A seamless transition
Schools can begin with a short pilot programme lasting one or two weeks before deciding on full-scale adoption. Once approved, implementation typically takes less than a week.
Teacher training has also been deliberately simplified. Recognising that not all educators are comfortable with advanced technology, the platform features an intuitive interface requiring minimal technical expertise.
Initial training usually takes only one to two hours, after which teachers can begin using the system independently. The company also provides ongoing support and optional review sessions to ensure schools continue using the platform effectively.
A new learning dawn
India’s National Education Policy (NEP) 2020 encourages schools to move beyond rote learning by promoting competency-based assessment, experiential learning and personalised education.
The ClassMap platform supports case-based and data-based questions that encourage analytical thinking rather than memorisation. It also enables personalised adaptive learning by identifying each student’s weakest areas and generating targeted practice material.
Looking ahead, the company plans to introduce customised learning content that will allow students to strengthen weaker concepts directly through the platform.
Building confidence in change
Introducing artificial intelligence into schools has not been without challenges. One of the biggest concerns among teachers was the fear that AI might eventually replace them. Convincing educators that the technology is intended to support rather than substitute classroom teaching required considerable effort.
The company consistently emphasises that ClassMap functions as an assistant rather than a replacement. By automating repetitive administrative tasks, it enables teachers to devote more time to teaching, mentoring and engaging with students. As schools have gained first-hand experience with the platform, this initial scepticism has gradually diminished, leading to wider acceptance of AI-assisted assessment.
From schools to horizons
Although ClassMap was initially developed for K–12 education, its underlying technology is applicable across a much broader educational landscape. The company is already working with coaching centres, Chartered Accountancy institutions, law universities and engineering programmes. Because the platform focuses on evaluating handwritten responses rather than a specific curriculum, it can be adapted to a wide range of academic disciplines. International expansion also forms part of the company’s long-term vision.
Contact details
Gaurav Jain
CEO & Co-Founder, Funscholar Innovations Private Limited, Kolkata, West Bengal
E: gaurav.jain@classmap.in