Gradia AI: Smart Grading, Smarter Learning

Abstract

In today’s rapidly evolving education landscape, educators—especially in underserved and resource-limited communities—face increasing challenges in providing timely and personalized feedback due to large class sizes and heavy workloads. Traditional manual grading methods are time-consuming and often inconsistent, hindering individualized support that is vital for student growth. This project, GRADIA AI – An AI-Powered Teacher Assistant, addresses these challenges by automating the grading process and generating personalized feedback for students. The system leverages Artificial Intelligence (AI) technologies such as Natural Language Processing (NLP), Machine Learning, and Computer Vision to evaluate both handwritten and digital student submissions. It provides detailed, constructive feedback aligned with each learner’s needs, enabling teachers to focus more on mentorship and less on repetitive evaluation tasks. The solution integrates multiple tools and technologies to ensure efficiency and scalability. The frontend is developed using HTML, CSS, and JavaScript, enabling an interactive user interface for teachers and students. The backend is powered by Python Flask, a lightweight web framework used for managing requests, APIs, and database interactions. Gemini AI API and Google Vision API are employed for intelligent grading, feedback generation, and text extraction from scanned exam papers. MongoDB serves as the primary database for storing user details, assignments, grades, and feedback, while Google Cloud Services provide reliable hosting, storage, and AI processing capabilities.

Introduction

Education systems are increasingly strained by larger class sizes, higher workloads, and the demand for personalized learning, particularly in underserved communities. Traditional manual grading is time-consuming, repetitive, inconsistent, and delays feedback, which negatively impacts student learning. To address these challenges, GRADIA AI is proposed as an AI-powered teacher assistant that automates grading and generates personalized feedback using Natural Language Processing (NLP), Machine Learning (ML), and Computer Vision (CV).

GRADIA AI processes handwritten and digital submissions through OCR (Google Vision API), evaluates answers via AI models (Gemini AI API), and generates constructive, individualized feedback. Teachers can review and approve AI-generated comments, maintaining human oversight. The platform is built with a modular architecture:

• Frontend: HTML, CSS, JS dashboards for teachers and students.

• Backend: Python Flask for server logic and AI API integration.

• Database: MongoDB for secure storage of submissions, grades, and feedback.

• Cloud deployment: Google Cloud for scalability and performance.

The system automates grading, feedback generation, and analytics, improving efficiency, consistency, and fairness while allowing teachers to focus on mentoring. It aligns with UN SDG 4 to ensure quality education through technology.

Objectives: Automate grading, generate personalized feedback, reduce teacher workload, provide intuitive dashboards, integrate OCR and AI evaluation models, ensure secure data storage, and enable scalable, fair, and accurate assessments.

Literature Survey: Existing AI-based educational tools demonstrate OCR, NLP, and deep learning can enhance grading, feedback quality, and performance tracking. However, challenges remain in scalability, personalized feedback, and combining text and image evaluation.

System Analysis: Current manual grading is labor-intensive, inconsistent, and delayed, limiting personalized learning. Schools—especially in resource-constrained regions—struggle with scalability, data-driven insights, and adaptive teaching strategies. GRADIA AI addresses these gaps by providing an automated, intelligent, and inclusive evaluation platform.

References

[1] A. Verma and R. Nair, “AI-Based Evaluation System for Automated Grading and Feedback in Education,” IEEE International Conference on Artificial Intelligence and Data Engineering (AIDE), 2023. [2] P. Singh and K. Patel, “Integration of Machine Learning Models for Automated Student Assessment,” International Journal of Advanced Computer Science and Applications (IJACSA), vol. 14, no. 3, 2024. [3] M. Gupta and S. Rao, “Design and Implementation of an Intelligent Teacher Assistant Using Natural Language Processing,” International Conference on Computational Intelligence and Smart Systems (CISS), 2022. [4] N. Sharma and L. Das, “AI-Powered Feedback Generation Using NLP Techniques,” Journal of Emerging Technologies in Education and Artificial Intelligence, vol. 12, no. 2, 2023. [5] R. Kumar and A. Thomas, “Automated Evaluation of Student Assignments Using Deep Learning,” IEEE International Conference on Data Science and Engineering (ICDSE), 2022. [6] S. Mehta and J. George, “Intelligent Academic Assistant for Teachers Using GPT and OCR Technologies,” International Journal of Innovative Research in Computer Science and Technology (IJIRCST), vol. 13, no. 4, 2024. [7] B. Reddy and M. Roy, “AI-Driven Assessment and Performance Analytics System for Educational Institutions,” International Conference on Artificial Intelligence and Education Technologies (AIET), 2021. [8] D. Sharma and C. Joseph, “Deep Learning-Based Student Performance Prediction and Automated Evaluation,” Journal of Computer Applications and Educational Technologies, vol. 11, no. 1, 2023. [9] K. Banerjee and R. Singh, “Enhancing Educational Quality Through AI-Based Grading and Personalized Feedback Systems,” IEEE International Conference on Smart Learning Environments (ICSLE), 2023. [10] T. Iyer and P. Dutta, “A Hybrid Framework for Automating Assessment Using NLP and Computer Vision,” International Journal of Artificial Intelligence Research, vol. 15, no. 2, 2024. [11] L. Menon and A. Jacob, “Teacher Support Systems Using AI for Performance Tracking and Feedback,” Proceedings of the International Conference on Intelligent Systems in Education (ISIE), 2022. [12] V. Chandra and M. Kapoor, “AI-Based Student Evaluation Using Optical Character Recognition and Text Analytics,” IEEE Conference on Computational Intelligence and Applications (ICCIA), 2023. [13] P. Nair and D. Kumar, “Design of a Scalable AI Grading System Using Flask and PostgreSQL,” International Journal of Intelligent Information Systems, vol. 10, no. 3, 2024. [14] S. Raj and R. Pillai, “Automation of Academic Evaluation Using Deep Neural Networks,” IEEE Transactions on Learning Technologies, vol. 17, no. 1, 2023. [15] J. Thomas and K. Fernandez, “AI in Education: Automating Assessment and Feedback Delivery,” International Journal of Educational Technology and AI Integration, vol. 9, no. 2, 2023. [16] A. Ghosh and P. Varma, “Intelligent Dashboard Systems for Teachers Using Machine Learning Analytics,” Proceedings of the Global Conference on Data Science and Artificial Intelligence (GCDSAI), 2024. [17] M. Ramesh and S. Iqbal, “Leveraging BERT and GPT Models for Context-Aware Academic Feedback,” Journal of Artificial Intelligence in Education Research, vol. 14, no. 1, 2024. [18] V. Patel and L. Menon, “Integration of AI and Cloud-Based Systems for Automated Assignment Evaluation,” IEEE Conference on Cloud Computing in Education (CCE), 2023.

Copyright

Copyright © 2025 Jesilyn L, Kavipriya B, Mr. S. Vinoth Kumar. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.