AI-powered pitch training platform with multi-agent mentoring, practice, and feedback system.
PitchQuest is an educational simulation platform that helps students practice business pitches through conversational AI agents. Based on research from "AI Agents and Education: Simulated Practice at Scale" by Mollick et al., it provides personalized mentoring, realistic investor interactions, and comprehensive feedback. The system uses three specialized AI agents orchestrated through LangGraph to create a complete learning experience.
- 🤖 Three Specialized AI Agents - Mentor, Investor, and Evaluator agents for complete training
- 💬 Interactive Chat Interface - Clean, responsive web UI for natural conversations
- 📊 Session Persistence - Save progress and continue across sessions
- 🔄 Structured Learning Flow - Guided journey from preparation to practice to feedback
- ☁️ Production Deployed - Serverless backend on AWS Lambda, frontend on Vercel
- 🎓 Research-Based Design - Pedagogically grounded in educational best practices
| Layer | Technology | Purpose |
|---|---|---|
| Frontend | Next.js 15.4, React, TypeScript | Modern web interface |
| Backend | FastAPI, AWS Lambda, API Gateway | Serverless API |
| Database | Supabase (PostgreSQL) | Session & message storage |
| AI/ML | OpenAI GPT-5-mini, LangGraph | Agent orchestration |
| Deployment | Vercel (Frontend), AWS (Backend) | Cloud hosting |
- Python 3.11+
- Node.js 18+
- OpenAI API key
- Supabase account (or local PostgreSQL)
# 1. Clone repository
git clone https://github.com/msalmancodes/PitchQuest.git
cd PitchQuest
# 2. Backend setup
python -m venv pitchquest_env
source pitchquest_env/bin/activate # Windows: pitchquest_env\Scripts\activate
pip install -r requirements.txt
cp .env.example .env # Add your API keys
uvicorn pitchquest_api.main:app --reload
# 3. Frontend setup (new terminal)
cd frontend
npm install
npm run dev
# 4. Open http://localhost:3000PitchQuest/
├── 📄 README.md # You are here
├── 📄 Developer_Onboarding.md # Complete setup guide
├── 📄 PROGRESS.md # Development timeline
├── 📄 requirements.txt # Python dependencies
├── 📄 config.py # Configuration settings
├── 📂 pitchquest_api/
│ ├── main.py # FastAPI application
│ ├── lambda_handler.py # AWS Lambda handler
│ └── routers/ # API endpoints
├── 📂 agents/
│ ├── mentor_agent.py # Mentoring logic
│ ├── investor_agent.py # Pitch session logic
│ └── evaluator_agent.py # Feedback generation
├── 📂 prompts/
│ └── *.yaml # Agent prompt templates
├── 📂 frontend/
│ ├── src/ # React components
│ └── package.json # Node dependencies
└── 📂 deployment/
└── build.sh # Lambda build script
- Developer Onboarding - Detailed setup, deployment, and API documentation
- Progress & Roadmap - Development history and future plans
- API Documentation - Endpoint specifications
- Agent Architecture - Multi-agent system design
- Muhammad Salman - Project Lead & Developer
- Claude (Anthropic) - Architecture design, code review, documentation
- ChatGPT (OpenAI) - Initial development, debugging support
- Mollick et al. - "AI Agents and Education" research paper
- Web Page: (https://pitch-quest-sj48.vercel.app/)
- API Endpoint:
https://am0h8n8b8i.execute-api.us-east-1.amazonaws.com/default/api/health - Frontend: [Contact for access]
MIT License - see LICENSE file for details
- Research: Based on "AI Agents and Education: Simulated Practice at Scale" by Ethan & Lilach Mollick et al.
- Technologies: OpenAI, AWS, Vercel, Supabase, FastAPI, Next.js, LangGraph
- Support: Wharton School's Generative AI Lab for the foundational research
Building the future of education through AI-powered practice 🎓