Stable Route AI

Applied AI prototype for selecting stablecoin transfer routes across different blockchain networks.

The system compares available transfer options by fee, delay, depeg risk, compliance constraints, and network conditions. It includes a backend API, a frontend interface, and optional LLM-based route explanation using a Mistral-based model.

Main idea

The project focuses on route selection for stablecoin transfers.

Given several possible routes, the system evaluates them and returns a ranked recommendation based on practical factors such as:

• transfer fee
• expected delay
• network
• selected token
• depeg risk
• compliance constraints
• route availability

Main features

• Backend API for route analysis
• Frontend interface for route selection
• Stablecoin and network comparison
• Compliance rule checking
• Real-time provider logic
• Optional Mistral-based explanation layer
• Docker-based deployment structure

Technical stack

• Python
• FastAPI-style backend structure
• React / Vite frontend
• Docker / Docker Compose
• Web3-related utilities
• Mistral-based local LLM inference
• Rule-based compliance layer

Repository structure

• backend/ — API routes, schemas, services, compliance logic
• frontend/ — React/Vite user interface
• docker-compose.yml — container orchestration
• Dockerfile — backend container setup
• requirements.txt — backend Python dependencies
• .env.example — example environment variables without real secrets

Security and data note

Real environment variables, API tokens, local model cache, logs, and audit files are not included in this public repository.

Use .env.example as a template for local configuration.

Status

Cleaned public version of an applied AI prototype.