Farhad

Nazari

Building production-grade quantitative and Web3 systems for crypto markets. KEY DELIVERABLES: ▸ Crypto Market Regime Detection System HMM + GMM + Change Point Detection · 11+ years BTC data Live interactive dashboard · Full ML pipeline from ingestion to deployment ▸ Crypto-FX Anomaly Detection Framework 4-method fusion (Isolation Forest, Z-Score, CUSUM, Volatility-Aware) Production-ready · Designed for stress conditions and regime shifts ▸ DeFi Token Swap Interface Full DEX with Uniswap V2 · React + Ethers.js + Solidity Live demo available ▸ TokenVault — dApp Rescue Case Study Diagnosed and rebuilt a broken DeFi application React + TypeScript + Solidity · Full case study documented (PDF) ▸ Portfolio Risk Analysis Engine VaR, CVaR, Monte Carlo simulations · Streamlit dashboard Deployed as live service ▸ FX Algorithmic Trading System Python + Polars + XGBoost · 20+ custom indicators Walk-forward backtesting with realistic slippage modeling Stack: Python · Polars · XGBoost · Scikit-learn

Applied machine learning with focus on financial time-series, anomaly detection, and production ML pipelines. ▸ Built anomaly detection systems using Isolation Forest, CUSUM, and ensemble methods ▸ Developed ML pipelines for financial sequential data: regime detection, change-point analysis, volatility modeling ▸ Designed feature engineering workflows for time-series financial data ▸ Implemented backtesting and model evaluation frameworks ▸ Delivered production-ready Python code with documentation Stack: Python · Pandas · Scikit-learn · Statsmodels · SQL