Syed Ammar Iqbal@syedammariqbal96

This project is an autonomous, multi-agent AI trading system designed to generate risk-adjusted P&L using Alpaca’s paper trading environment. Developed solo over a 7-day sprint, the system orchestrates a Regime Agent to classify daily market conditions and three specialized Execution Agents that trade distinct options strategies: Iron Condors on SPY, Calendar Spreads on tech names, and Bull Put Spreads on high-beta underlyings. True to agentic design principles, every strategic decision—from regime classification and strike selection to entry timing—is executed via an LLM through a structured tool-calling loop. However, capital preservation is strictly enforced by a deterministic, hard-coded Gatekeeper that validates every proposed trade against 9 strict risk rules (position sizing, drawdown limits, DTE constraints) before submission. The LLM never judges risk limits. The infrastructure features a resilient multi-provider LLM router with automatic fallback chains (Groq, OpenRouter, Gemini), daily token budgeting, and SQLite state persistence. Currently running live via VS Code for over 36 hours on a fresh $100k paper account (PA30BDXWDF0M), the system has successfully navigated real-world friction—including LLM rate limits, partial multi-leg fills, and JSON parsing failures—to execute profitable NVDA bull put spreads.

🌎 Online Hackathon | 💻 Dates: 28 August–4 September 2026 | 🏆 $6,000 Prize Pool Build AI trading agents on Alpaca — autonomous agents and trading apps using Alpaca's Trading API, MCP server and CLI. 🌍 Fully online hackathon. Join and build from anywhere in the world.