An intelligent AI Note-Generator that extracts, summarizes, translates, and speaks back notes from audio — powered by Flask, Gemini API, and AI-powered ETL (Extract–Transform–Load) processing.

EduHacks AI Note-Generator is a Flask-based backend application that automates lecture and meeting note-taking. You simply upload an audio file, and it performs a full AI-powered ETL pipeline:

1. 🎤 Transcription — Converts speech to text.
2. 🧠 Summarization — Generates bullet-point summaries using LexRank.
3. 🗂 Flashcards — Creates intelligent Q&A flashcards.
4. 🔊 Text-to-Speech (TTS) — Converts summaries back to audio.
5. 🌍 Translation — Translates summaries using the Gemini API into your preferred language (default: Urdu 🇵🇰).

git clone https:https://github.com/ck-ahmad/EduHacks_AI_Note_Creator.git
cd eduHacks-ai-note-taker

python -m venv .venv
source .venv/bin/activate      # Windows: .venv\Scripts\activate

pip install -r requirements.txt

Create a .env file (or use environment variable):

API_Key_2=YOUR_GEMINI_API_KEY

You can get your API key from:
🔗 https://aistudio.google.com/app/apikey

python -c "import nltk; nltk.download('punkt'); nltk.download('punkt_tab')"

python app.py

Server runs on 👉 http://localhost:5000

GET /health

Response:

{
  "status": "ok",
  "message": "EduHacks AI Note-Taker backend running"
}

POST /api/upload

Form Data:

• file: audio file (.mp3, .wav, .m4a, etc.)
• lang: optional target language code (default "ur" for Urdu)

Response:

{
  "transcript": "...",
  "bullets": ["...", "..."],
  "flashcards": [
    {"question": "...", "answer": "..."}
  ],
  "translated_summary": "...",
  "files": {
    "transcript_txt": "/outputs/xxxx_transcript.txt",
    "summary_txt": "/outputs/xxxx_summary.txt",
    "flashcards_txt": "/outputs/xxxx_flashcards.txt",
    "translated_summary_txt": "/outputs/xxxx_summary_translated_ur.txt",
    "audio_summary": "/outputs/xxxx_summary.mp3"
  }
}

Files are saved in the /outputs folder:

• transcript.txt → raw speech-to-text output
• summary.txt → bullet summary
• flashcards.txt → generated Q&A pairs
• summary_translated_ur.txt → translated summary
• summary.mp3 → AI-generated audio version

You can access them directly via:

GET /outputs/<filename>

• .wav
• .mp3
• .m4a
• .aac
• .ogg

The backend uses Gemini 2.5 Flash via the official google-genai SDK.

Example (translator.py):

from google import genai

class GeminiTranslator:
    def __init__(self, api_key, model="gemini-2.5-flash"):
        self.client = genai.Client(api_key=api_key)
        self.model = model

    def translate(self, text, target_language):
        prompt = f"Translate the following text to {target_language}:\n\n{text}\n\nOnly return translation."
        response = self.client.models.generate_content(model=self.model, contents=prompt)
        return response.text.strip()

📦 eduHacks-ai-note-taker
│
├── app.py
├── translator.py
├── speech_to_text.py
├── summarizer.py
├── flashcards.py
├── tts.py
│
├── uploads/            # Uploaded audio files
├── outputs/            # Generated files (txt, mp3)
├── templates/
│   └── home.html       # Optional frontend
├── static/
│
├── requirements.txt
└── README.md

flask
flask-cors
werkzeug
pydub
SpeechRecognition
sumy
nltk
pyttsx3
gTTS
pydantic==2.8.2
google-genai

• 🔹 Multi-language UI (frontend translation toggle)
• 🔹 Database support (store notes, summaries, and metadata)
• 🔹 Real-time transcription (WebSocket)
• 🔹 User authentication & dashboards
• 🔹 Audio segmentation for longer files
• 🔹 Front-end React or Next.js integration

This project is licensed under the MIT License — free for educational and personal use.