Edunova

Revolutionizing the Education space

Comment

Edunova: Revolutionizing Education with Agentic AI

Inspiration

As students, the founders of Edunova observed that educators dedicate an average of five hours weekly, totaling over 140 hours annually, to repetitive tasks such as grading and evaluation. This time, they noted, could be better spent on teaching, mentoring, and inspiring students. They also identified inconsistencies in grading standards among teachers, leading to unequal evaluations, and a lack of personalized feedback for individual student needs. Witnessing the rapid advancements of AI in education, the Edunova team was inspired to reimagine assessments as an enabler rather than a burden. Edunova was founded to return valuable time to educators while ensuring fairness, accuracy, and personalization in evaluations.

What it Does

Edunova is an Agentic AI-powered education automation platform that integrates assessment, question generation, and analytics into a seamless workflow. The platform utilizes multiple specialized AI agents to achieve its functionalities:

  • Question Generation Agent: This agent dynamically creates diverse question sets tailored to various subjects, chapters, and difficulty levels.
  • Evaluation Agent: The Evaluation Agent is responsible for grading a variety of response formats, including text, diagrams, code, and structured answers, through an agentic pipeline.
  • Feedback Agent: This agent personalizes student insights by pinpointing individual strengths and weaknesses, offering tailored recommendations for growth and development.

These agents collectively orchestrate a closed-loop automation workflow: Create → Evaluate → Analyze → Feedback, which aims to deliver efficiency, fairness, accuracy, and actionable insights.

How We Built It

The development of Edunova commenced with surveys conducted among educators to understand their daily workflows and identify critical pain points. The team then utilized IBM’s Agent Development Kit (ADK) to design modular AI agents with specialized roles, enabling them to collaborate within an orchestrated workflow. IBM’s Granite models power the LLM-based grading and generation components, while data preprocessing toolkits from IBM help normalize multi-format student responses, such as text, images, and code. Each agent is containerized for reusability and can be deployed across diverse institutional settings, including urban universities and rural schools with limited resources.

Challenges We Ran Into

A primary challenge in developing Edunova was building trust in automation among educators, who often expressed concerns regarding credibility, bias, and transparency in fully agent-driven grading. The Edunova team addressed this by incorporating explainable AI outputs, which illustrate the reasoning behind a particular grade, and by implementing adaptive evaluation criteria that institutions can customize. Another significant challenge was handling multimodal inputs, as diagrams, programming code, and descriptive answers require specialized preprocessing. IBM’s data toolkits were leveraged to standardize these diverse inputs, ensuring consistent evaluation.

Accomplishments That We're Proud Of

Edunova has achieved significant accomplishments, including:

  • Time Efficiency: The platform has reduced grading time from hours to mere seconds, saving educators an average of 25 hours per month, which can be reinvested into teaching and mentoring.
  • Scalable Question Generation: Educators can now generate an entire exam paper in under 30 seconds, a task that previously took hours of manual preparation. The platform also maintains full control over subject, topic, and difficulty customization.

What We Learned

The Edunova team discovered that agentic AI is particularly well-suited to address India-centric educational challenges, such as high teacher workloads, large student-teacher ratios, and diverse learning needs. This makes it a natural fit for reinventing the classroom in the AI era.

What's Next for Edunova

Edunova is currently focused on scaling its platform through a credibility and integration framework that includes:

  • Audit Logs & Traceability: Implementing audit logs and traceability for every agent decision to enhance transparency and accountability.
  • Alignment with National Education Policy (NEP 2020): Integrating Edunova into formal curricula by aligning with the National Education Policy (NEP 2020).
  • Multi-lingual Support Agents: Developing multi-lingual support agents to enable AI-powered assessments across India’s diverse regional languages.

Edunova's vision is to establish itself as a blueprint for India’s AI-augmented education framework, demonstrating how agentic AI can deliver real-world impact at scale.

Built With

Share this project:

Updates