The Rez backend is a modular Python-based system designed for high reliability and intelligent trade execution. It is divided into three primary functional engines:
The AI-driven core. It uses a ReAct pattern to orchestrate multi-turn reasoning cycles with Large Language Models, transforming indicators into trade decisions.
The computational layer. It calculates technical indicators like EMA, RSI, and MACD locally using Binance data and pandas-ta, ensuring the agent has access to low-latency, accurate signals.
