🚀 QuantX: AI-Powered Quantitative Investment & Portfolio Orchestration Platform

QuantX is a production-grade, microservice-based quantitative trading, backtesting, and portfolio optimization platform. It combines deep learning forecasting models (LSTMs, GRUs, Transformers), reinforcement learning decision engines (PPO), real-world data ingestion pipelines, and a high-fidelity, real-time React/Next.js dashboard built with dark-mode glassmorphism aesthetics.

---

🏗️ System Architecture & Microservices

QuantX is engineered with a modular, resilient microservices architecture communicating via a central API Gateway:

graph TD
    GW[API Gateway - Port 8005] --> |Routes Requests| MD[Market Data Service - Port 8001]
    GW --> |Routes Requests| FE[Feature Service - Port 8002]
    GW --> |Routes Requests| SG[Signal & Decision Service - Port 8003]
    GW --> |Routes Requests| PM[Portfolio & Execution Service - Port 8004]
    GW --> |Routes Requests| AI[AI Prediction Service - Port 8006]
    GW --> |Routes Requests| QR[Quantum Research Service - Port 8007]
    
    FE --> |Reads Features| DB[(SQLite Local DB)]
    MD --> |Seeds Prices| DB
    PM --> |Fetches Targets & Rebalances| DB
    SG --> |Evaluates RL Rules| DB
    AI --> |LSTM/GRU Inference| DB

Loading

In production, the API Gateway acts as a unified hub with a built-in SQLite persistent fallback layer. If the downstream microservices are offline (common on lightweight cloud hosting), the Gateway executes orders and rebalancing algorithms locally against quantx_local.db using the exact same quantitative rules.

---

✨ Key Features

1. AI Predictions & Inference Engines:
   • Deep learning forecasting models (LSTM, GRU, Transformers) for stock prices.
   • Reinforcement learning (PPO) agents executing real-time buy/sell/hold decisions.
2. Real-world Database Ingestion:
   • Automated startup seeder populating the SQLite database with 60+ assets (Nifty 50, NASDAQ, Crypto) and 30 days of historical price history.
3. High-Performance Backtesting Lab:
   • Code-editor interface executing historical backtests with live metric charting (Sharpe, Drawdown, Profitability).
4. Automated Portfolio Optimization:
   • Markowitz mean-variance optimization and automatic asset rebalancing console utilizing SciPy solvers.
5. Real-time UI Over WebSockets:
   • Real-time ticker price feeds, system alerts, and notification drawer.

---

📂 Project Structure

QuantX/
├── .github/workflows/          # GitHub Actions CI/CD workflows
├── backend/
│   ├── api-gateway/            # central router & WebSocket server
│   ├── market-data-service/    # historical stock price ingestion
│   ├── portfolio-service/      # order manager & Mean-Variance optimizer
│   ├── ai-prediction-service/  # LSTM, GRU, and Transformer price predictors
│   ├── signal-service/         # DRL PPO agent execution
│   ├── backtesting-service/    # vector strategy simulation engine
│   ├── quantum-research-service/# advanced strategy research console
│   └── populate_db.py          # database seeder
├── frontend/
│   └── dashboard/              # Next.js React Dashboard application
├── requirements.txt            # Python dependencies (for local/CI)
├── requirements-render.txt     # lightweight dependencies (for Render deployment)
├── start_services.py           # local development runner script
└── README.md                   # this README

---

🛠️ Tech Stack

• Frontend: Next.js 14, React 18, Recharts, Lucide Icons, TailwindCSS.
• Backend Services: Python (FastAPI, Uvicorn, SQLAlchemy).
• AI & Machine Learning: PyTorch, Stable Baselines 3 (RL), Pandas, NumPy.
• Database & Persistence: SQLite (local development / offline mode), PostgreSQL (production-ready schemas).

---

🚀 Getting Started (Local Setup)

Prerequisites

• Python 3.10+
• Node.js 18+

1. Backend Installation & Database Setup

Initialize a virtual environment and install backend dependencies:

# Create and activate virtual environment
python -m venv .venv
.venv\Scripts\activate     # On Windows
source .venv/bin/activate  # On macOS/Linux

# Install dependencies
pip install -r requirements.txt

# Seed the database with Nasdaq & Nifty 50 datasets
python backend/populate_db.py

2. Frontend Setup

Install frontend node dependencies:

cd frontend/dashboard
npm install
npm run build   # Compile production package

3. Launching Services

You can run all services with a single startup script:

# From the root directory:
python start_services.py

This automatically starts the API Gateway, Market Data, Feature, Portfolio, Signal, AI Prediction, and Next.js web application servers. Open http://localhost:3000 in your browser to interact with the platform!

---

🧪 Running Tests

QuantX includes an extensive suite of unit and system tests ensuring service reliability:

# Run unit tests
pytest tests/

---

📊 Endpoints Cheat-Sheet

Service	Port	Endpoint	Description
API Gateway	8005	GET /api/health	Service status aggregator
API Gateway	8005	GET /api/portfolio	Fetch active holdings & positions
API Gateway	8005	POST /api/trade	Route orders (w/ local fallback)
API Gateway	8005	POST /api/portfolio/rebalance	Trigger portfolio optimizer rebalance
API Gateway	8005	POST /api/backtest	Trigger custom strategy backtests
Market Data	8001	GET /assets	List registered stock assets
AI Prediction	8006	GET /predict/{symbol}	Get price forecasting scores

---

🗺️ Future Roadmap

• Live Broker Integration: Connect the Execution Engine to broker APIs (e.g., Alpaca, Interactive Brokers, Zerodha) to transition from paper trading to live trading.
• Advanced Sentiment Agent: Stream news feeds and financial statements to extract real-time market sentiment via pre-trained LLMs.
• Decentralized Deployments: Distribute and run the microservices in separate geographical containers using Kubernetes or Docker Swarm.
• Enhanced ML Backtesting: Expand backtesting to include transaction cost models, execution slippage, and historical borrow costs for short sales.