EDUGEN-AI PLATFORM
An AI-powered multilingual learning platform that transforms educational content into notes, quizzes, videos, and more.
Created on 23rd April 2025
•
EDUGEN-AI PLATFORM
An AI-powered multilingual learning platform that transforms educational content into notes, quizzes, videos, and more.
The problem EDUGEN-AI PLATFORM solves
Edu-Gen Platform is designed to revolutionize how students and educators interact with educational materials. By integrating cutting-edge AI technologies, it automates several time-consuming learning tasks, making the educational process smarter, faster, and more personalized.
Content to AI Video: Transforms textual content into engaging, narrated educational videos with visual elements—perfect for remote learning or flipped classrooms. Video to PDF: Extracts text from educational videos using YOLO and OCR, converting them into detailed, searchable PDFs. Content Summarization: Summarizes lengthy documents or articles using Retrieval-Augmented Generation (RAG), providing quick and relevant insights. MCQ Generation: Instantly generates customizable multiple-choice questions from any PDF using the BART model, simplifying quiz and exam preparation. Notes Generation: Converts dense documents into concise English notes, leveraging Mixtral-8x and B-32768 models. Multilingual Support: Summarizes content and answers questions in multiple languages, making education accessible to a diverse set of learners. Memorization Evaluation: Evaluates spoken content against study material and provides grades based on accuracy using Llama and GroQ.
In essence, Edu-Gen empowers both students and educators by streamlining learning and content creation in a seamless, multilingual environment.
Challenges I ran into
One of the main challenges was achieving reliable accuracy in OCR for different video types and lighting conditions. We tackled this by fine-tuning YOLO for slide detection and applying adaptive pre-processing filters before text extraction.
Handling multilingual content was another big hurdle, particularly ensuring that summaries and Q&A generation in non-English languages remained coherent and accurate. We overcame this by integrating GroQ acceleration for fast inference and using fine-tuned versions of RAG and Llama for different languages.
Performance optimization was also crucial, especially during video processing and AI video generation, which required balancing quality with computation time.
Tracks Applied (1)
Groq track
Groq
Technologies used
Cheer Project
Cheering for a project means supporting a project you like with as little as 0.0025 ETH. Right now, you can Cheer using ETH on Arbitrum, Optimism and Base.