Inspiration

With the rapid growth of AI tools like Chat-GPT and other generative models, traditional exams are becoming easier to bypass using AI assistance. We were inspired to build a system that ensures fair evaluation by designing questions that test deep understanding, critical thinking, and real-world application, rather than simple memorization.

What it does

AI-Proof Pareeksha is an intelligent exam generation engine that:

Creates AI-resistant questions (case-based, analytical, subjective) Detects pattern-based answers often generated by AI Generates dynamic question sets for each student Supports multiple difficulty levels (JEE, UPSC, etc.) Encourages conceptual clarity over rote learning

How we built it

We used a combination of:

Frontend: HTML, CSS (Neon-themed UI for modern look) AI Logic: NLP-based question generation and validation

Workflow:

Input topic + difficulty AI generates multiple question types Engine filters out “AI-easy” questions Final paper is generated dynamically

Challenges we ran into

Designing questions that are truly AI-proof Avoiding repetitive or predictable question patterns Balancing difficulty vs fairness Handling AI hallucinations in generated content

Accomplishments that we're proud of

Built a working prototype in limited hackathon time Successfully generated high-quality analytical questions Designed a clean and interactive UI Created a system that promotes honest learning

What we learned

Importance of prompt engineering in AI systems How AI can be both a problem and a solution Real-world challenges in education technology Working with APIs, backend logic, and UI integration

What's next for AI-Proof Pareeksha: Exam Design Engine

Add AI answer detection system Integrate student performance analytics Build a live exam platform Add voice-based and adaptive questioning Expand to school, college, and competitive exams

Share this project:

Updates