AlphaAgent
Autonomous AI agent that paper-trades stocks, crypto & indices using multi-agent debate and real-time market analysis
Built at AgentX Hackathon 2026
Team NullPointer: Deepansh Gupta, Pranay Nair, Tanvi Kumbhar, Saloni kori
Built with Next.js, TypeScript, Groq AI, Alpaca API, TailwindCSS
About the project
AlphaAgent is an AI-powered paper trading system that autonomously analyzes markets and executes trades. It uses a three-agent debate system (Technical Analyst, Risk Officer, Skeptic) powered by Groq's LLM to make BUY/SELL/HOLD decisions on 8 assets including Indian stocks (RELIANCE, HDFCBANK, TCS, NIFTY 50), US tech (NVDA, AAPL), and crypto (BTC, ETH).
Key Features:
• Multi-agent AI debate system for trading decisions
• Real-time news sentiment analysis
• Backtesting with Momentum & Mean Reversion strategies
• Risk management with daily loss limits & position sizing
• Alpaca Paper Trading API integration
• Real-time portfolio tracking & performance charts
• Autonomous 60-second trading cycles
Tech Stack: Next.js, TypeScript, Groq AI (gpt-oss-120b), Alpaca API, TailwindCSS