DeepMentor_demo.zip(Demo Video of the application)

DeepMentor 🧠

DeepMentor is a powerful, interactive learning platform designed to help students understand complex mathematical concepts. It combines a knowledgeable AI tutor with an automated animation engine to provide a rich, multi-modal learning experience.

Features

• AI-Powered Study Assistant: Get detailed, step-by-step explanations for textbook questions. The AI tutor's knowledge is strictly limited to the provided textbook and solution manual, ensuring accurate and relevant answers.
• Automated Mathematical Animations: Visualize complex mathematical concepts with automatically generated Manim animations. The system takes a mathematical explanation and creates a corresponding video to illustrate the logic.
• Interactive Chat Interface: Engage with the AI tutor through an intuitive, chat-like interface built with Streamlit.
• RAG-Powered Responses: The AI leverages a Retrieval-Augmented Generation (RAG) pipeline with Pinecone's vector database to pull in the most relevant information from the source material.
• Extensible and Modular: The codebase is organized into logical modules, making it easy to extend or modify.

Architecture

The application is built on a modular architecture that combines several key technologies:

• Frontend: A user-friendly web interface created with Streamlit (app.py).
• AI and Language Models: The core logic for generating explanations and animation plans is powered by the Groq API (codegen.py), which provides fast inference for large language models.
• Animation Engine: Mathematical animations are generated using the Manim Community Edition library. The codegen.py script produces Manim code, which is then executed by manim_runner.py.
• Retrieval-Augmented Generation (RAG): The AI's responses are grounded in the provided educational materials using a RAG pipeline.
• Vector Database: Pinecone is used to store and retrieve vector embeddings of the textbook content, ensuring that the AI can quickly find the most relevant information to answer a user's query (vectordb.py, retriever.py).
• PDF Content Extraction: The extract.py script processes PDF documents, converting them into a format that can be used to build the vector database.

File Descriptions

File	Description
app.py	The main Streamlit application. It handles the user interface, chat history, and orchestrates calls to the other modules.
codegen.py	Contains the AnimationGenerator class, which uses the Groq API to generate animation plans and Manim code.
manim_runner.py	A utility to execute Manim code and find the path to the most recently rendered video.
retriever.py	Implements the RAG pipeline. It uses Pinecone and LangChain to retrieve relevant document chunks and generate a response.
vectordb.py	A script to create and store vector embeddings from the source PDFs into the Pinecone database.
extract.py	A utility for extracting content from PDF files, preparing it for the vector database.
asset.py	Stores prompt templates and other assets used by the AI models.
requirements.txt	A list of all the Python dependencies required to run the project.
Dockerfile	A Dockerfile to containerize the application and its dependencies for easy deployment.
.env	A file to store environment variables, such as API keys.

Setup and Installation

1. Clone the repository:

   git clone https://github.com/muvva-krishna/manim.git
   cd DeepMentor
   

2. Install system dependencies: The application requires FFmpeg and a LaTeX distribution. These are included in the Docker setup.

3. Install Python dependencies:

   pip install -r requirements.txt

4. Set up environment variables: Create a .env file in the root directory and add your API keys:

   GROQ_API_KEY="your_groq_api_key"
   LANGCHAIN_API_KEY="your_langchain_api_key"
   PINECONE_API_KEY="your_pinecone_api_key"
   OPENAI_API_KEY="your_openai_api_key"
   

5. Prepare the Vector Database: Run the vectordb.py script to populate your Pinecone index with the content from the provided PDF(s).

   python vectordb.py

How to Run

To start the application, run the following command in your terminal:

streamlit run app.py

This will launch the Streamlit web server, and you can access the application in your browser at http://localhost:8501.

Dependencies

• Streamlit: For the web interface.
• Groq: For fast language model inference.
• Manim: For creating mathematical animations.
• OpenAI: For generating embeddings.
• LangChain: For building the RAG pipeline.
• Pinecone: For the vector database.
• python-dotenv: For managing environment variables.
• pdf2image: For extracting content from PDFs.