REAL – Resilient Environments for Adversarial Learning

💡 Inspiration

We wanted to make cyber warfare training more realistic and more accessible. Existing platforms often feel rigid, require extensive setup, or don’t let participants fully control the network environment. We set out to create a tool that combines a cyber range with a training management system, giving users the power to design networks, run red-team/blue-team exercises, and track performance — all in one place.

🎯 What it does

REAL lets teams:

• Design custom networks with drag-and-drop nodes (servers, firewalls, endpoints).
  
• Spin up environments on demand, automatically provisioning containers for realistic training scenarios.
  
• Run exercises and track results, with built-in scoring, logging, and reporting.
  
• Collaborate in real time, allowing multiple participants to edit the network or respond to incidents together.
  

🛠️ How we built it

• Frontend: Built with Next.js for a responsive, fast-loading interface and SSR capabilities.
  
• Backend: A FastAPI service handles network generation, user management, and scoring logic.
  
• Database: Supabase provides authentication, row-level security, and real-time updates for collaboration.
  
• Infrastructure: Used container orchestration to dynamically create and destroy training environments.
  
• UI/UX: Focused on making network design intuitive with a visual editor and live updates for all users in the exercise.
  

⚠️ Challenges we ran into

• Network Simulation Fidelity: Balancing realism with performance so environments spin up quickly.
  
• Security Isolation: Making sure offensive tools could be run safely in a sandboxed environment.
  
• Real-Time Sync: Handling concurrency issues when multiple users modified the same network topology.
  
• Load Management: Orchestrating multiple simultaneous training environments without over-provisioning.
  

🏆 Accomplishments that we're proud of

• Successfully created a working end-to-end app that goes from network design to live cyber exercise.
  
• Built a real-time collaborative editor using Supabase subscriptions.
  
• Achieved a smooth user experience with Next.js that makes complex network design approachable.
  
• Ensured secure isolation between exercises to prevent cross-environment interference.
  

📚 What we learned

• How to build a full-stack application with Next.js, Supabase, and FastAPI working in sync.
  
• Best practices for secure container orchestration and sandboxing of potentially malicious code.
  
• Importance of clear user flows in technical training tools — good UX makes cyber exercises less intimidating.
  
• How to optimize real-time data sync for multiple collaborators without losing consistency.
  

🔮 What's next for REAL

• Advanced analytics: Adding dashboards to visualize performance trends over multiple exercises.
  
• Scenario marketplace: Letting instructors share and import pre-built network scenarios.
  
• AI-powered adversaries: Introducing automated red-team bots to challenge defenders dynamically.
  
• Scaling infrastructure: Supporting larger, more complex network topologies for enterprise-level training.
  

Built With

• nextjs
• supabase