All projects

TheArena – AI-Powered Trading Signals Platform

Role: Founder Completed
TheArena – AI-Powered Trading Signals Platform
Role
Founder
Status
Completed
Technologies
20 Used

What It Is

TheArena is a complete AI trading signals platform built from the ground up. Traders subscribe to receive AI-generated buy/sell signals for crypto and forex markets, with real-time notifications, an economic calendar, a live community chat room, and performance analytics. The platform runs across three separate servers — a Django backend, a React frontend, and a standalone FastAPI/FreqAI ML microservice — all containerized and deployed at thearena.cloud.

The Problem

Retail traders either pay for expensive signal services with no transparency into how signals are generated, or they do all their own analysis manually. There was no affordable platform that combined AI-generated signals with clear reasoning, a tiered access model, and a real-time community — so I built one.

How I Solved It

The Django backend handles authentication (including TOTP-based 2FA), tiered subscription management, payments via NOWPayments (USDT/crypto), signal delivery with tier-based access control, and real-time chat via Django Channels and WebSockets. Signals are delivered to traders via browser push notifications and Telegram — with free tier users receiving signals on a 10-minute delay and paid tiers getting them instantly. The React + TypeScript frontend (Vite, Tailwind CSS v4, Zustand) covers the full trader journey — onboarding, dashboard, signals, economic calendar, settings, and payment flows. The FreqAI microservice is a separate FastAPI server with its own PostgreSQL database. It fetches OHLCV market data via CCXT and Yahoo Finance, engineers 17+ technical indicators, trains RandomForest and GradientBoosting models, and serves buy/sell/hold predictions with confidence scores. Django calls this service to generate and validate signals, with Redis caching predictions for fast delivery.

The Impact

  • Live at thearena.cloud — a real deployed product
  • 3 servers, 10 running Docker services (web, worker, beat, frontend, nginx, FreqAI server, 2x PostgreSQL, 2x Redis)
  • Fully automated signal pipeline from market data → ML model → trader notification
  • Tiered access model with free and paid tiers, real payment processing

Tech Stack

Google Sheets API React Twelve Data TypeScript FastAPI Django Channels Vite Tailwind CSS Zustand React Hook Form Zod scikit-learn CCXT Yahoo Finance API Celery Celery Beat WebSockets Docker NOWPayments Telegram API

Demo

No demo video available yet. Check back soon.