NoteCrew

Description

This app, hosted in Streamlit, uses CrewAI agents to structure information into easy-to-review notes. The inputs for the app are text, PDF, or DOCX files.

Problem Statement

Like many college students, I have notes scattered across Google Docs, PDFs, Notion pages, and even screenshots of notes I was too lazy to re-copy. Now, you can extract the text using screenshots, as it has become a cool feature(wondering how? It uses OCR (Optical Character Recognition)).

But once the text is extracted, arranging it, fact-checking that you wrote the correct statement, and formatting it into clear, structured notes is still time-consuming. Instead of doing that manually, why not let a team of AI agents handle all that work?

Features:

• Multi-Format Input Support: Process notes from PDFs, Google Docs, plain text
• Agent -Powered Note Structuring: Automatically formats your notes into popular study structures: Cornell, Outline, or Boxing method
• Grammar EnhancementL: Identifies and corrects grammar issues for improved clarity
• Fact-Checking: validate key facts in your notes for accuracy
• Optional Flashcard Generation: Converts your content into Q&A-style flashcards to help with active recall and better studying.

Information Flow

• If the input is not direct text, the app extracts it from the uploaded file.
• The extracted or pasted text is passed to a crew of AI agents.
• The agents process the input sequentially, each performing a distinct part in preparing the final structured notes.

Agents

• Grammar Agent: Corrects grammar and writing inconsistencies in the input.
• Fact-check Agent: Validates the factual accuracy of the content.
• Format Agent: Applies a selected note-taking method:
  • Outline Method
  • Cornell Method
  • Boxing Method
• (Optional) Flashcards Agent: Creates 5 to 15 flashcards summarizing key facts and concepts from the input text.

Technologies

---

Installing

1. Prerequisites

• Python >=3.10 and <3.13 must be installed.
• This project uses uv for dependency and package management.

2. Install uv

pip install uv

3. Clone the Repository

git clone https://github.com/Rigo-stack/crewai-note.git
cd crewai-note

4. Install Dependencies

uv pip install -r requirements.txt

5. Set Up Environment Variables

OPENAI_API_KEY=your_openai_key
SERPER_API_KEY=your_serper_key

6. Launch App:

Streamlit run app.py

---

Usage:

Use the Interface:

• Upload your notes as text, PDF, or DOCX
• Choose a note-taking method (Cornell, Outline, or Boxing)
• Optionally enable flashcard generation
• Click Run to let the AI agents process and format your notes

Customization of Agents:

• Modify 'agents.yaml' to define the agents
• Modify 'tasks.yaml' to define the tasks
• Modify 'crew.py' to add your own logic, tools, and specific args
• Modify 'app.py' to add custom inputs for your agents and tasks

---

Contributing:

Pull requests are welcome. For major changes, please open an issue first to discuss what you would like to change. Please make sure to update tests as appropriate

---

License

This project is licensed under the MIT License.

---