
AI Day-Trader Development for Multi-Account Strategies
Upwork
Remoto
•10 hours ago
•No application
About
Build Multi-Account AI Day-Trader (Political Mirror | Short Scalper | Long Swing) Objective Develop an AI trading system that autonomously manages three brokerage accounts, each following a distinct, rule-based strategy with live risk controls and data transparency. 1. POL-TRACK (Account A) Tracks and ranks disclosed politician trades (STOCK Act / QuiverQuant / CapitolTrades). Mirrors only proven outperformers based on win-rate and risk-adjusted return. Executes identical long/short entries with strict stop-loss + position-size rules. 2. SHORT-SCALP (Account B) Shorts equities showing parabolic extensions or overbought signals. Executes rapid intraday “sell to open, buy to close” trades on liquid tickers. Includes borrow check, SSR awareness, OCO orders, and end-of-day flatten. 3. LONG-SWING (Account C) Goes long on fundamentally strong or technical breakout stocks. Multi-day/week holds: “buy to open, sell to close.” Adds trailing stops, pyramiding logic, and overnight-risk management. Deliverables Three autonomous AI agents (each tied to separate broker APIs – IBKR/Alpaca). Data-scraper modules for political filings + market feeds. Unified dashboard (P&L, exposure, Sharpe, drawdown). Risk engine (max loss per symbol/day, kill switch). Dockerized deployment + backtest suite + docs. Required Skills Python (Backtrader/vectorbt), broker APIs, ML forecasting, short-sale mechanics, compliance (SEC/FINRA).