I need a production-ready, AI-assisted trading automation system written in Python that plugs directly into Alpaca Markets, IBKR / TWS, and TradeZero. The core service will run on a VPS (Ubuntu preferred) inside Docker, expose its endpoints through FastAPI, and receive strategy signals from TradingView webhooks. Once a webhook arrives, the service should route orders to the correct broker, apply position-sizing logic, set stop-loss / take-profit brackets, and check real-time volatility to ensure the trade fits predefined risk limits. The system must: • Authenticate safely with each broker using encrypted key storage and automatic rotation. • Offer a toggle between paper trading, live trading, and historical back-testing, sharing one code path so results stay consistent. • Log every request, response, fill, error, and balance change to PostgreSQL, then surface summaries on a lightweight dashboard; alerts go out instantly via Telegram and email. • Provide an admin API for strategy configuration, portfolio caps, and emergency kill-switches. • Support three strategy styles out of the box—day trading, swing trading, and scalping—while leaving hooks for future strategies. Technical stack requirements: Python 3.11+, FastAPI, Claude API for signal enhancement, Docker, PostgreSQL, GitHub Actions for CI/CD, plus standard back-testing libraries (vectorbt / backtrader or similar). Clear docstrings, a concise README, and inline comments are mandatory so another engineer can pick up the repo with minimal ramp-up. Acceptance criteria 1. End-to-end trade flows execute correctly on Alpaca, IBKR/TWS, and TradeZero in both paper and live modes. 2. Position sizing, stop-loss / take-profit, and volatility checks block any order that would breach limits. 3. Historical runs reproduce the same fills given the same data set. 4. Deployment scripts spin up the entire stack on a fresh Ubuntu VPS in one command. 5. All secrets remain outside the repo and are mounted securely at runtime.