Skip to content

beepboop2025/LiquiFi

Repository files navigation

LiquiFi

Autonomous treasury management platform with real-time FX monitoring, ML-driven forecasting, and automated hedging execution.

LiquiFi connects to 12 global data sources, forecasts liquidity using LSTM/GRU/Transformer ensemble models, and automates regulatory compliance -- all from a single glassmorphism dashboard with smooth animations and WebSocket-driven live updates.


Features

  • Real-Time FX Rate Monitoring -- Live WebSocket ticker for USD/INR, EUR/INR, GBP/INR, JPY/INR, and 30+ additional money market fields scraped from RBI, CCIL, FBIL, FRED, ECB, and more.
  • Automated Hedging Execution Engine -- Order lifecycle management with a circuit breaker that prevents runaway execution during volatile markets, including half-open probe limiting to avoid oscillation.
  • ML Ensemble Forecasting -- LSTM, GRU, and Transformer models trained on 24 engineered features with walk-forward cross-validation. 24-hour forecast horizon with weighted ensemble inference.
  • Monte Carlo Stress Testing -- 300-path simulation with VaR and CVaR computation for tail risk analysis.
  • Risk Management Dashboard -- Instrument exposure visualization, stress test scenarios, and a compliance engine covering CRR, SLR, ALM gap analysis, LCR, and NSFR.
  • Portfolio Analytics and P&L Tracking -- Cash position waterfall, yield attribution, and active alert management.
  • Backend Simulation Engine -- Full order lifecycle simulation with rate drift modeling and priority-based rate merging from multiple data sources.
  • Regulatory Automation -- CRR/SLR position tracking, ALM gap analysis, LCR/NSFR computation, and branch-level reporting.
  • Desktop Application -- Ships as a native macOS app via Electron with the Python backend bundled as an extra resource.

Tech Stack

Layer Technology
Frontend React 18, TypeScript, Vite 5
Visualization Recharts, Lucide React
Desktop Electron 33, electron-builder
Backend FastAPI, Python 3.12+
Machine Learning LSTM, GRU, Transformer (custom), Monte Carlo simulation
Database PostgreSQL 15, Redis 7, Alembic migrations
Real-Time WebSocket (server push + client subscription)
Infrastructure Docker Compose

Getting Started

Prerequisites

  • Node.js 18+
  • Python 3.12+
  • Docker (for PostgreSQL and Redis)

Frontend

npm install
npm run dev

The development server starts at http://localhost:5173.

Backend

cd backend
python -m venv venv
source venv/bin/activate
pip install -r requirements.txt

# Start PostgreSQL and Redis
docker compose -f docker-compose.dev.yml up -d

# Run database migrations
alembic upgrade head

# Start the server
python main.py

The API server starts at http://localhost:8000 with WebSocket support on /ws.

Desktop (Electron)

# Development
npm run electron:dev

# Production build (macOS)
npm run electron:build

License

MIT

About

Treasury Automation Platform — real-time money market data, LSTM/GRU/Transformer forecasting, PostgreSQL backend

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors