Skip to content

axeljuan02/ARES-Infrastructure

Repository files navigation

ARES — Event-Driven Earnings Intelligence Homelab

ARES is a personal, always-on finance homelab designed to analyze corporate earnings events through a structured, institutional-style workflow.

Built at the intersection of quantitative finance, event-driven market research, and infrastructure engineering, ARES aims to replicate—at a small, educational scale—the analytical discipline used by professional buy-side desks during earnings season.

ARES is not a trading bot and does not execute orders.
Its purpose is to deliver clear, explainable, and statistically grounded earnings intelligence.


🎯 Project Objective

ARES was developed with three main objectives:

  • Understand earnings-driven price dynamics
    By studying EPS surprises and historical post-earnings price reactions.

  • Build a production-inspired research system
    Running continuously on a personal server, with disciplined data handling and reproducible logic.

  • Develop institutional market intuition
    By transforming raw earnings data into concise, desk-style briefings supported by historical evidence.

This project is part of a long-term academic and professional trajectory toward quantitative trading and systematic research.


⚙️ Processing Pipeline

main_pipeline earnings_pipeline mailing_pipeline

🖥️ Infrastructure

ARES runs continuously on a lightweight personal homelab:

  • Raspberry Pi 5 (Pironman 5 case)
  • Headless Linux environment
  • Python-based services
  • JSON for structured storage
  • Event-driven execution aligned with earnings calendars
  • Local LLM-8850 accelerator (TBD)
1 2

🛤️ Roadmap

Module Status
Raspberry Pi construction ✅ Done
Event-driven orchestration (main.py) ✅ Done
Persistent state & idempotency ✅ Done
Earnings ingestion (FMP API) ✅ Done
Event study engine (post-earnings returns) ✅ Done
EPS surprise segmentation & hit-rate statistics ✅ Done
Interpretative earnings summary ✅ Done
Automated HTML email reporting ✅ Done
Inline visualizations (heatmap) ✅ Done
Raspberry Pi 24/7 deployment ✅ Done
Local LLM-assisted synthesis (LLM 8850 accelerator + Pytorch) 🚧 Planned

📁 About This Repository

ARES is a personal research and learning project intended to:

  • explore event-driven equity behavior,
  • practice building production-style financial systems,
  • document institutional-grade analytical reasoning.

It makes no claim of predictive certainty and is not intended for live trading.


👤 Author

Axel Juan
BBA Third-Year Student — ESSEC Business School
CFA Level I Candidate
Aspiring Quantitative Trader

ARES is part of a broader personal initiative combining finance, programming, and systems engineering.

🚧 Actively developed — scope and features will evolve.

About

ARES is a miniature, always-on market intelligence system inspired by hedge fund research infrastructure. Designed as a 24/7 micro-server, it continuously monitors U.S. equity markets, processes earnings releases, and delivers automated, institution-style insights.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors