Skip to content

An AI-powered educational content transformation system that converts static documents into structured, interactive multimedia lessons.

Notifications You must be signed in to change notification settings

3lis0/TeachFlow

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

56 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

TeachFlow

The Future of Education - AI-Powered Course Content Generation

🎯 Vision

To revolutionize the way instructors design and deliver education by harnessing the power of Generative AI to create personalized, engaging, and high-quality learning experiences at scale.

🚀 Mission

  • Automate the creation of complete course materials—from curriculum structure to slides and video scripts
  • Empower instructors with AI tools that reduce content preparation time and increase teaching impact
  • Enhance learning quality through well-structured, narrative-rich lessons tailored to different age groups and learning styles
  • Bridge the gap between raw online content and classroom-ready resources with intelligent content transformation

📊 Problem Statement

Current Challenges in Education Content Creation:

  • ⏰ Time-consuming: Searching for reliable content across the internet takes hours
  • 🔄 Manual effort: Summarizing, structuring, and designing slides is repetitive and slow
  • 📉 Inconsistent quality: Content varies between instructors and lacks standardization
  • 📈 Low scalability: Difficult to quickly adapt materials for different levels, age groups, or learning styles
  • 🎥 Video creation barriers: Requires extra skills and effort, often beyond the instructor's capacity

✨ Key Features

🗺️ Intelligent Learning Roadmap Creation

Builds adaptive roadmaps from uploaded PDFs, structuring course content into modules, lessons, and topics with hierarchical organization.

🌐 Language and Teaching Style Adaptation

Customizes content based on the learner's preferred language (English, Arabic, and mixed) and instructional style for a personalized experience.

📊 Content Extraction to PPTX

Automatically converts static PDF materials into well-structured, presentation-ready PowerPoint slides with professional formatting.

🎬 Narrative and Video Script Generation

Generates clear, engaging narrative text and video scripts tailored for multimedia lesson delivery and enhanced student engagement.

Processing Pipeline:

  1. Upload Documents → PDF files uploaded to the system
  2. Pages to Images → Convert PDF pages to image format
  3. OCR Processor → Extract text using Optical Character Recognition
  4. Text Patches → Break text into manageable chunks
  5. Text Merger → Combine and structure text coherently
  6. Content Structure → Organize into modules, lessons, and topics
  7. JSON Converter → Convert to structured JSON format
  8. UI Display → Present hierarchical structure to users

🛠️ Technology Stack

Core Technologies

Technology Purpose Version
Google Gemini 2.5 Flash Large Language Model for content generation Latest
CrewAI Autonomous Agent Framework for task orchestration Latest
Google Cloud Services Scalable cloud infrastructure -
Firebase Real-time backend and database (Firestore) Latest
Flask Lightweight web framework for REST APIs 2.0+
OCR Technology Text extraction from PDF documents -

Key Technology Benefits

  • 🤖 Gemini 2.5 Flash: Fast and cost-effective understanding, summarization, and generation capabilities
  • 🔄 CrewAI: Modular orchestration of task-specific AI agents for automated workflows
  • ☁️ Google Cloud: Scalable, secure, and seamless integration with AI tools
  • 🔥 Firebase: Real-time database, authentication, and analytics support
  • ⚡ Flask: Quick REST API development and AI workflow integration

🚀 Getting Started

Prerequisites

Python >= 3.8
Flask >= 2.0
Firebase Account
Google Cloud Platform Account
CrewAI Framework
OCR Libraries

Installation

# Clone the repository
git clone https://github.com/adelian14/TeachFlow.git
cd TeachFlow

# Create virtual environment
python -m venv teachflow_env
source teachflow_env/bin/activate  # On Windows: teachflow_env\Scripts\activate

# Install dependencies
pip install -r requirements.txt

# Set up environment variables
cp .env.example .env
# Configure your API keys and settings in .env

# Initialize Firebase
firebase login
firebase init

# Run the application
python app.py

Environment Configuration

# Google AI Configuration
GOOGLE_API_KEY=your_gemini_api_key_here
GOOGLE_CLOUD_PROJECT_ID=your_project_id

# Firebase Configuration
FIREBASE_API_KEY=your_firebase_api_key
FIREBASE_AUTH_DOMAIN=your_project.firebaseapp.com
FIREBASE_DATABASE_URL=https://your_project.firebaseio.com
FIREBASE_PROJECT_ID=your_project_id
FIREBASE_STORAGE_BUCKET=your_project.appspot.com

# CrewAI Configuration
CREWAI_API_KEY=your_crewai_key

# Application Settings
FLASK_ENV=development
FLASK_DEBUG=True
UPLOAD_FOLDER=uploads/
MAX_CONTENT_LENGTH=50MB

🎯 Target Customers

Primary Markets

  1. Independent Instructors & Trainers - Freelance educators and corporate trainers
  2. Universities & Academic Institutions - Higher education content creation
  3. Online Learning Platforms - E-learning content providers
  4. Vocational Training Centers - Professional skill development organizations
  5. Government & Educational Ministries - Public education content standardization

💰 Business Model

Revenue Streams

  1. B2C SaaS Subscriptions - Individual instructor plans
  2. B2B Institutional Licensing - Enterprise and academic licenses
  3. Custom AI Integration & Services - Tailored solutions for large organizations
  4. Content Marketplace Commission - Future feature for content sharing
  5. Assessment & Quiz Generator - Premium add-on features

Cost Structure

  • Development and maintenance costs
  • AI & Cloud infrastructure expenses
  • API licensing fees (Google, CrewAI)
  • Marketing and customer acquisition
  • Customer support operations
  • Legal compliance and data security

📈 SWOT Analysis

Strengths ✅

  • Solves fundamental time-saving problem in education
  • Modular workflow with independent AI agents
  • Multi-language support (Arabic, English, mixed)
  • Comprehensive feature set (roadmaps, slides, scripts)
  • Flexible teaching style customization
  • Scalable SaaS architecture
  • Strategic educational partnerships

Weaknesses ⚠️

  • Heavy reliance on input data quality
  • No human review component
  • Accuracy limitations in specialized fields
  • Current processing time constraints

Opportunities 🚀

  • Rapid e-learning market growth
  • LMS integration potential (Moodle, Blackboard)
  • Automated assessment features
  • Arabic-focused AI tool advantage
  • Platform partnerships (Udemy, Coursera)

Threats ⚡

  • Competition from established AI education tools
  • User concerns about AI-only content generation
  • Data privacy and security requirements
  • API dependency risks
  • Institutional resistance to AI adoption

🔮 Future Roadmap

Phase 1: Core Enhancement

  • Video Lesson Generation - Automated video content creation
  • Automated Assessments & Quiz Generator - Intelligent testing tools
  • Performance Optimization - Reduce processing times
  • Quality Assurance Tools - Content review mechanisms

Phase 2: Content Expansion

  • Multimodal Content Integration - Support for various media types
  • Multiple File Format Support - Beyond PDFs (DOCX, PPTX, etc.)
  • Advanced Language Support - Enhanced Arabic and multilingual capabilities
  • Custom Templates - Industry-specific content templates

Phase 3: Platform Integration

  • LMS Integrations - Moodle, Blackboard, Canvas compatibility
  • API Marketplace - Third-party integration ecosystem
  • Mobile Applications - iOS and Android apps
  • Collaborative Features - Team-based content creation

🔧 Current Limitations & Challenges

Processing Performance

  • Issue: Relatively long processing times for complex documents
  • Impact: User experience and scalability concerns
  • Mitigation: Ongoing optimization and infrastructure improvements

Quality Assurance

  • Issue: Generated content may not always align with instructor goals
  • Impact: Manual editing often required
  • Mitigation: Developing review mechanisms and customization options

Language Support

  • Issue: Currently optimized for English; Arabic support in development
  • Impact: Limited market reach in Arabic-speaking regions
  • Mitigation: Prioritizing Arabic language model improvements

Cost Management

  • Issue: Dependency on third-party LLM APIs creates recurring costs
  • Impact: Operational expense scaling with usage
  • Mitigation: Exploring cost optimization and alternative models

🔐 Security & Privacy

Data Protection

  • All uploaded documents are encrypted in transit and at rest
  • User data is processed according to GDPR and educational privacy standards
  • Document content is not permanently stored without user consent
  • Access controls ensure institutional data privacy

AI Ethics

  • Transparent AI-generated content labeling
  • Bias detection and mitigation in content generation
  • Human oversight recommendations for critical educational content
  • Opt-out options for data usage in model improvements

👥 Team Members

📄 License

This project is licensed


TeachFlow: Let instructors teach — and TeachFlow does the rest 🎓🤖✨

About

An AI-powered educational content transformation system that converts static documents into structured, interactive multimedia lessons.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 54.4%
  • HTML 45.6%