🎬 Manima

Agentic AI for Curriculum-Aligned Educational Video Generation

Manima is a web application that enables educators to generate curriculum-aligned educational videos from simple text prompts or uploaded study materials (PDFs). It uses a sophisticated Agentic AI pipeline powered by LangGraph to orchestrate multiple specialized agents that plan, script, code, and review animations using Manim.

---

🚀 Overview

The platform automates the creation of high-quality educational content by simulating a production studio:

• Planner Agent: Breaks down topics into structured learning segments.
• Script Agent: Writes clear, engaging narration and visual descriptions.
• CodeGen Agent: Generates Python code for Manim animations.
• Reviewer Agent: Validates code and ensures quality.
• Engine: Executes code in a sandboxed Docker environment to render videos.

---

🧩 Tech Stack

🧠 AI & Brain

• LangGraph: For stateful, multi-agent orchestration.
• LangChain: For LLM interactions.
• LLMs: LLaMA 3.1 70B / Mistral 8x7B (via Groq/OpenRouter).
• RAG: Supabase Vector (pgvector) + local embeddings.

⚙️ Backend (The Engine)

• FastAPI: REST API for orchestration and job management.
• Python 3.10+: Core language.
• Manim: Mathematical Animation Engine.
• Docker: Sandboxed execution environment for rendering.
• FFmpeg: Video stitching and post-processing.
• Supabase: Database (PostgreSQL) and Storage.

🖥️ Frontend

• Next.js: React framework for the UI.
• Tailwind CSS: Styling and responsive design.
• Supabase Auth: User authentication.

---

🏗️ Architecture

The backend follows a 3-layer architecture:

flowchart LR
    User[User] -->|Prompt/PDF| API[FastAPI Layer]
    API -->|Trigger| Brain[LangGraph Brain]
    
    subgraph Agents
        Planner --> Scripter
        Scripter --> Reviewer
        Reviewer --> Coder
    end
    
    Brain --> Agents
    Coder -->|Manim Code| Engine[Docker Engine]
    Engine -->|Video Segments| Storage[Supabase Storage]
    Storage -->|Final Video| User

Loading

---

🛠️ Project Setup

For a detailed, step-by-step setup guide for Mac and Windows, please refer to SETUP.md.

Quick Start (Mac/Linux)

1. Prerequisites

• Node.js & pnpm
• Python 3.10+
• Docker Desktop
• Supabase CLI

2. Clone the Repo

git clone https://github.com/mustansirr/AnimEd-AI.git
cd AnimEd-AI

3. Start Database

supabase start

4. Frontend Setup

cd frontend
pnpm install
# Configure .env.local with Supabase URL & Anon Key
pnpm dev

5. Backend Setup

cd backend
python -m venv venv
source venv/bin/activate
pip install -r requirements.txt
# Configure .env with Supabase URL & Service Role Key
uvicorn main:app --reload

---

🧾 License

This project is open-source under the MIT License.