-
Notifications
You must be signed in to change notification settings - Fork 0
FAQ
AI Education Pilot is an open-source, AI-powered educational analytics platform that helps teachers track student progress, generate intelligent insights, and provide personalized learning experiences.
Yes! AI Education Pilot is 100% free and open-source under the MIT License. However, you'll need your own OpenAI API key for AI features, which has its own costs.
- Frontend: Next.js 15 (React, TypeScript, Tailwind CSS)
- Backend: Python (FastAPI)
- Database: PostgreSQL
Absolutely! AI Education Pilot is designed to be self-hosted. See our Installation Guide for details.
- Sign in to your account
- Click "Create Module" from the home dashboard
- Enter module name and description
- Click "Create"
- Share the access code with students
Students can join in two ways:
-
Access Code: Give them the 6-digit code, they visit
/join -
Direct Link: Share
your-domain.com/join/ACCESS_CODE
Supported formats:
- Documents: PDF, DOCX, DOC, TXT
- Presentations: PPTX, PPT
- Images: JPG, JPEG, PNG, GIF
- Archives: ZIP, RAR
- Multiple Choice
- True/False
- Short Answer
- Essay/Long Answer
- Fill in the Blank (coming soon)
- Matching (coming soon)
When students submit answers, AI analyzes their responses against the correct answer and rubric (if provided) to generate personalized, constructive feedback.
Yes! Navigate to Dashboard → Rubric to create custom rubrics with:
- Multiple criteria
- Point values
- Descriptions
- AI-powered rubric generation
- Individual student performance
- Class-wide statistics
- Progress tracking over time
- Question difficulty analysis
- Engagement metrics
- Performance trends
Minimum:
- 2 CPU cores
- 4 GB RAM
- 10 GB storage
Recommended:
- 4+ CPU cores
- 8+ GB RAM
- 20+ GB SSD storage
See Installation Guide for details.
Yes, for AI features like:
- AI-generated feedback
- Test question suggestions
- Analytics insights
- Document Q&A (RAG)
Basic features work without an API key.
PostgreSQL 15+ is the primary database. It stores:
- User accounts
- Modules and tests
- Student responses
- Analytics data
Currently, the system is built for OpenAI's API. However, you can modify the code to support:
- Anthropic Claude
- Google PaLM
- Open-source models (LLaMA, Mistral)
Not yet, but the web interface is fully responsive and works on:
- Mobile browsers
- Tablets
- Desktop
- Passwords are hashed with bcrypt
- JWT tokens for authentication
- HTTPS encryption (in production)
- Role-based access control
- Database encryption at rest
No. Your data stays on your server. OpenAI processes text for AI features but doesn't store it (per their policy).
Yes! You can export:
- Student lists (CSV, JSON)
- Test results (CSV, JSON)
- Analytics reports (CSV, PDF)
- All module data
The system provides the tools for FERPA compliance, but you're responsible for:
- Proper deployment security
- Access controls
- Data handling policies
- Privacy notices
There's no hard limit. Performance depends on your server resources. Typical deployments handle:
- Small: 1-100 students
- Medium: 100-1,000 students
- Large: 1,000+ students (requires scaling)
Yes! Each teacher has their own account and modules. Students can be enrolled in multiple modules.
Database backup:
pg_dump ai_pilot > backup.sqlRestore:
psql ai_pilot < backup.sqlSee Backup Guide for details.
We're working on import tools. Currently, you can:
- Manually recreate modules
- Import student lists via CSV
- Import questions (with custom scripts)
Check:
- Access code is correct (6 digits)
- Module is active
- Server is running
- No firewall blocking access
Verify:
- OpenAI API key is set in
.env - API key is valid and has credits
- Backend server is running
- Check backend logs for errors
Common causes:
- No student data yet
- Database connection issues
- Frontend can't reach backend API
- Check browser console for errors
Solutions:
- Check file size limit (default 10MB)
- Verify file format is supported
- Ensure upload directory has write permissions
- Check backend logs
- Fork the repository
- Create a feature branch
- Make your changes
- Submit a pull request
See Contributing Guide.
Yes! The frontend uses:
- Tailwind CSS for styling
- Shadcn/ui components
- Next.js for routing
Customize in Frontend/app/ and Frontend/components/.
Free:
- AI Education Pilot software
- Self-hosting
Paid (you provide):
- Server/hosting
- OpenAI API credits
- Domain name (optional)
- SSL certificate (optional, Let's Encrypt is free)
Depends on usage. Typical costs:
- GPT-4: ~$0.03 per 1K tokens
- GPT-3.5: ~$0.002 per 1K tokens
Estimate: $0.01-0.10 per student feedback
Yes, if you:
- Self-host on existing hardware
- Don't use AI features (or use free alternatives)
- Use Let's Encrypt for SSL
Check our Video Tutorials page.
Yes! See the User Manual.
- Issues: GitHub Issues
- Questions: GitHub Discussions
- Chat: Community Discord (link in repo)
Ask in GitHub Discussions - we'll add it to the FAQ!
Still have questions? Ask the Community | Report Issues