Project

Notes2Quiz

Turning handwritten or photographed notes into something actually study-ready (summaries, audio, quizzes) usually takes several separate tools.

SERVICEAI Engineering & Full-Stack
YEAR2026
CREDITSFatin Israq Talha
LINK
Notes2Quiz Interface Showcase

Technology Architecture & Stack

PythonStreamlitGoogle GeminigTTS

Developed an all-in-one Streamlit study assistant. Users upload photos of handwritten or printed notes. The app leverages Google Gemini vision capabilities to extract handwritten text and synthesize concise study summaries. The summary is converted into downloadable audio narration via gTTS, and an interactive multiple-choice quiz (Easy, Medium, Hard difficulty levels) is generated for instant student self-assessment.

Key Features & Functionality

  • Vision OCR and handwriting comprehension using Google Gemini API.
  • Automated summary generation tailored for quick review.
  • Text-to-speech audio conversion via gTTS for auditory learning on the go.
  • Dynamic multiple-choice quiz generation with custom difficulty options and immediate grading.

System Architecture & Technical Design

Combines multimodal vision parsing with LLM structured prompt output parsing to build educational assessments in a single pass.

Key Engineering Takeaways

Designing multi-modal pipelines required strict prompt schema engineering to ensure consistent quiz JSON structures from raw OCR text.