ai-trading-crew

Multi-agent AI stock analysis coordinates specialized teams of agents to evaluate equities, assess risk, and deliver consensus trading decisions. This system uses a Head Coach, risk veto, and Devil's Advocate to reduce bias and manage complex workflows.

To set up an automated trading workflow, connect data feeds from Polygon for market analysis and use Alpaca for simulated execution in paper trading mode. Launch the AutoGen workflow to analyze US stock symbols and generate trading decisions.

Yes, you need data feeds from Polygon for market data and Alpaca for simulated paper trading execution. These integrations provide the necessary market context and execution environment for the agents to analyze equities and test strategies.

Reducing bias in automated stock analysis is achieved by implementing a Devil's Advocate agent and a risk veto mechanism. This multi-layer review process challenges consensus decisions before execution, ensuring robust risk assessment across the portfolio.

Yes, ChromaDB RAG memory enables cross-session knowledge retention and rapid retrieval of market context. This allows the multi-agent system to access historical analysis and maintain continuous strategy testing across live or simulated environments.

A 50-agent architecture requires significant coordination overhead and relies on external data feeds from Polygon and Alpaca. Without these connections, the system cannot perform stock analysis or execute simulated trades for the equity portfolio.