radare

Industrial data reconciliation platform using Lagrange Multipliers and chi-squared validation for mass and energy balance integrity.

Architecture

apps/
  backend/            Go API and reconciliation engine (Port 8080)
  webapp/             React SPA (Port 5173 dev / 80 prod)
database/
  coredb_migrations/  Core database SQL migrations
  logdb_migrations/   Audit log database SQL migrations
  coredb_seeds/       Initial seed data
  cmd/migrate/        Database migration CLI
client-api/           Bruno API collection
.gitea/workflows/     CI/CD pipeline definitions

Components

• apps/backend: Go backend utilizing GORM and gonum for mathematical optimization and balance solving.
• apps/webapp: React frontend built with Vite, TanStack Router/Query, and React Flow.

Reconciliation Engine

Process nodes contain measured streams and tags. The reconciliation algorithm applies Lagrange Multipliers to minimize weighted squared deviations while satisfying conservation constraints. Statistical integrity is evaluated via $\chi^2$ hypothesis testing.

Development

Prerequisites

• Go 1.25+
• Node.js 20+
• Docker / Podman

Running Services

Initialize databases, run migrations, and apply seeds:

make db-bootstrap

Start full development environment:

make dev

Database Management

make migrate        # Run coredb migrations
make seed           # Apply coredb seeds
make migrate-log    # Run logdb migrations
make seed-log       # Apply logdb seeds
make nuke-and-pave  # Reset database volumes and reapply migrations

Service Endpoints

Service	Type	Port	Endpoint
Web App	Frontend	5173	http://localhost:5173
HTTP API	Backend	8080	http://localhost:8080
PostgreSQL Core	Database	5432	localhost:5432/radare
PostgreSQL Logs	Database	5433	localhost:5433/radare_logs
Redis	Cache	6379	localhost:6379

Documentation

• 📋 Roadmap & TODOs - Planned features and project roadmap
• 📐 Architecture - System architecture and components
• 📄 License - GNU General Public License v3.0