NemoFish

<div align="center">

<img src="./static/image/NemoFish_logo_compressed.jpeg" alt="NemoFish Logo" width="65%"/>

### Swarm Intelligence Engine for Predictive Analytics

**Multi-agent reasoning · Consensus scoring · Autonomous execution**

[![GitHub Stars](https://img.shields.io/github/stars/ASGCompute/NemoFish?style=flat-square&color=DAA520)](https://github.com/ASGCompute/NemoFish/stargazers)
[![License](https://img.shields.io/badge/License-GPL--3.0-blue?style=flat-square)](./LICENSE)
[![ASG Compute](https://img.shields.io/badge/ASG%20Compute-R%26D-000?style=flat-square)](https://asgcompute.com)

[Quick Start](#-quick-start) · [How It Works](#-how-it-works) · [Dashboard](#-unified-control-room) · [Credits](#-credits)

</div>

---

## Overview

**NemoFish** is a swarm intelligence platform that coordinates multiple AI agents to generate probabilistic forecasts for competitive sports events. Each agent brings a different analytical lens — statistical modeling, psychological profiling, market signal analysis, scenario simulation — and they reach consensus through weighted voting. The system then scores opportunities by expected value and can execute positions on prediction markets autonomously.

This project began as a fork of [**MiroFish**](https://github.com/666ghj/MiroFish), an open-source multi-agent simulation engine developed by the CAMEL-AI research community. We extended MiroFish with a full-stack sports analytics pipeline and a unified control room for real-time monitoring.

> **An R&D project by [ASG Compute](https://asgcompute.com)** — exploring whether coordinated multi-agent reasoning can identify inefficiencies in prediction markets.

---

## 📊 Scale & Data Coverage

<div align="center">

| Metric | Value |
|:-------|------:|
| **Historical matches analyzed** | **971,320** |
| **Unique players profiled** | **33,962** |
| **Tournaments covered** | **9,201** |
| **Years of ATP data** | **1968 – 2026** (57 seasons) |
| **Surfaces modeled** | Hard · Clay · G