About

Summary: A "Decision-Grade" automated pipeline that transforms raw economic data from the Federal Reserve (FRED) into actionable market signals. This project demonstrates a complete production lifecycle: dynamic data ingestion, trend-following logic, and deployment via a professional FastAPI microservice.

🧠 Core Features:
Dynamic Data Ingestion: Automated connection to FRED fetching 2,500+ daily S&P 500 records. No static CSVs; the model adapts to the most recent market close.

- Modular Architecture: Built using Zerve’s DAG (Directed Acyclic Graph) structure, separating data handling, model logic, and performance validation.

- Backtested Strategy: Implements a 20-day trend-following SMA with an integrated "Auditor" block that calculates real-world win rates.

Production Deployment: Fully deployed via FastAPI with a live endpoint and interactive Swagger documentation.

🛠️ Tech Stack:
Language: Python (Pandas, Requests, Matplotlib)

Framework: FastAPI

Data Source: St. Louis Fed (FRED)

Infrastructure: Zerve Canvas & API Service

🔗 Live API Access:
For real-time programmatic access, please use the FastAPI endpoint provided in the links below.
Endpoint: https://market-logic-ai.hub.zerve.cloud/
Interactive Docs: https://market-logic-ai.hub.zerve.cloud/docs