"WesCodies team on Alpaca AI Trading Agents Hackathon Hackathon"

Team Idea

We are building an AI-powered behavioral trading agent that learns from a trader’s historical trades to identify recurring mistakes and cognitive biases, starting with confirmation bias. The system continuously analyzes new trades using fundamentals, technical indicators, market context, and the trader’s historical behavior. A multi-agent validation system challenges the trader’s thesis, generates evidence-based counter-theses, and debates conflicting conclusions until reaching consensus or identifying uncertainty. As the trader continues trading, the system learns from outcomes, updates their behavioral profile, and adapts its future analysis—creating a personalized AI trading co-pilot that helps traders recognize and correct repeated decision-making patterns.

Team Members

Sipho Chipanga

Sipho

Ngqabutho Sithole

Ngqabutho

Submission

Gypsi Avenger

Gypsi is an AI-powered trading platform with a conversational agent, real-time market analysis, deterministic trading logic, and MCP tools, enabling users to analyze markets, manage risk, and trade across platforms through one unified interface.