Anirban Pal@Anirban9372

AlgoSentinel is a fully autonomous AI-powered options trading agent built for the Alpaca AI Trading Agents Hackathon. Every 15 minutes during market hours, AlgoSentinel pulls live headlines from Google News RSS feeds focused on SPY and the S&P 500. These headlines are passed to Google Gemini 3.6 Flash, which performs real-time financial sentiment analysis and returns a structured signal: BULLISH, BEARISH, or NEUTRAL — along with a confidence score and reasoning. Before any trade is placed, the agent enforces 5 strict risk gates: 1. Equity floor — account must hold above $85,000 2. Maximum 3 simultaneous open positions 3. Per-trade budget capped at 5% of portfolio 4. Confidence threshold — signal must be ≥ 60% confident 5. Market hours enforcement — trades only 9:30 AM to 4:00 PM ET If all gates pass, the agent selects the closest ATM (at-the-money) SPY options contract using live Alpaca market data, calculates the optimal quantity within budget, and places a market order via the Alpaca Trading API — all in seconds. Once positions are open, a live monitor checks every cycle and automatically exits at 50% loss (stop-loss) or 100% gain (take-profit), protecting the portfolio without any human intervention. Every decision — whether a trade is placed or skipped — is logged to trades.log with full timestamp, signal, confidence, and reasoning for complete auditability. On September 4, 2026, AlgoSentinel fired its first real autonomous trade: 151 contracts of SPY 769 Put, deploying $4,983 based on a BEARISH signal driven by rising Treasury yields and overvaluation metrics. Tech stack: Python · Google Gemini 3.6 Flash · Alpaca Trading API · Google News RSS · paper trading on a $100k account.

🌎 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.