Team Idea
AlphaLoop Autonomous is a multi-agent trading system engineered for the Alpaca AI Trading Agents Hackathon. It pairs LLM reasoning with deterministic risk guardrails to execute rule-bounded options strategies in Alpaca's paper trading environment.
Architecture:
Market Agent: Queries live options chains, IV percentiles, and underlying price action using Alpaca’s Model Context Protocol (MCP) server.
Strategy Agent: Evaluates defined-risk plays (e.g., credit spreads, covered calls, delta-neutral setups) and generates step-by-step trade theses.
Risk Guardrail Layer: Deterministically validates orders against strict caps on portfolio Greeks, max per-trade risk, margin limits, and stop-loss rules before execution.
Execution & Telemetry Agent: Dispatches multi-leg orders via Alpaca’s Trading API and logs transparent reasoning traces to a live dashboard.
Tech Stack: Python, Alpaca MCP Server, Alpaca Trading API, NumPy/Polars, Streamlit/FastAPI.
Team Members
Shambhu Shekhar Sinha
shambhu_shekhar_sinha
