Freqtrade Crypto Bots
Overview
3 algorithmic trading strategies in dry run on OKX. EMA breakout for infrastructure tokens (SOL, AVAX), Donchian reclaims for momentum, elite scoring for meme coins (BONK, PEPE). Survived a 48% market crash with only -6% loss. Currently safe-good, not profit-good.
Three algorithmic trading strategies running in dry-run mode on OKX — no real money, but live market data.
EMA breakout strategy targets infrastructure tokens (SOL, AVAX) — riding momentum on confirmed breakouts above key moving averages.
Donchian channel reclaims target momentum coins — entering when price reclaims a key range boundary after a pullback.
An elite scoring system filters meme coins (BONK, PEPE) by volume, momentum, and relative strength before entering.
Survived a 48% market crash with only -6% drawdown. The bots are safe-good but not yet profit-good — risk management works, edge still needs tuning.
Stack
Backend / API
Tooling / AI
Infra / Deploy
Database
Other
Inspiration
Wanted to understand algorithmic trading properly — not just buy signals, but how strategies behave under real market stress.
Structure
Architecture
Schema
How it works
Strategy Design
Define entry/exit rules, timeframes, and indicator logic for each strategy.
Backtesting
Run historical data through Freqtrade's backtesting engine to validate strategy logic.
Hyperopt
Optimize strategy parameters using Freqtrade's hyperoptimization framework.
Dry Run
Deploy strategies in dry-run mode on OKX — live market data, no real capital.
Monitoring
Telegram bot sends trade notifications and daily performance summaries.
Analysis
Review trade logs, drawdown behavior, and edge quality after market events.
More projects
Kusina ni Lola
A family-owned Filipino restaurant. Showcasing a vibrant menu, online ordering, and reservation system, designed to bring the warmth of Lola's cooking to life.
AI Object Detection
My Cardio
A commission health monitoring web app focused on cardiac health. Users log their health data and receive AI-assisted and algorithm-based consultation feedback. Built with React and TypeScript on the frontend, Node.js and PostgreSQL on the backend. The consultation logic combines rule-based algorithms with AI-generated responses to give users meaningful health insights without replacing professional medical advice.
Myelin-core
Community knowledge base for low-level systems performance — optimizations, gotchas, and snippets for C, Zig, Rust, and ASM. Machine-first API designed for both humans and agents.