Usama B.

I'm a Top Rated Plus Quantitative Researcher and Developer on Upwork, currently completing my MS in Data Science at Harvard, with production experience building and shipping algorithmic trading systems across futures, options, and equities. My work spans the full pipeline: from signal research and rigorous backtesting through execution architecture and live broker deployment. Over the years I've collaborated with professional traders, systematic fund managers, and independent CTAs on strategy research, execution system builds, and quantitative infrastructure — working with live accounts and real capital, not toy portfolios. - Research Philosophy Most algorithmic trading research fails before it reaches live markets. The reasons are almost always the same: data snooping bias, fragile parameterization, fill simulation that doesn't reflect actual execution, and strategy logic never stress-tested against regime change. My research process is designed to make it hard to fool yourself. That means strict chronological train/test splits with no parameter bleedthrough, walk-forward optimization over anchored and rolling windows, block bootstrap for inference on autocorrelated returns, and parameter sweep frameworks built to expose fragility rather than maximize backtest metrics. A strategy that only works on one parameterization isn't a strategy, it's a coincidence. I care about robustness across parameter space, regime stability, and understanding exactly why a signal works before scaling it. I work extensively in QuantConnect's LEAN engine for research-grade backtesting and live account deployment, though deployment targets vary by strategy and client stack. IB TWS via ibapi, Tradovate, and NinjaTrader are all in regular rotation depending on execution requirements. I build custom analytics layers on top of backtest output: Sharpe decomposition, drawdown attribution, execution cost modeling, and signal decay analysis, because standard tearsheet metrics rarely tell the full story. - Strategy Coverage My research spans futures (NQ, ES), SPX and index options, ETFs, and crypto. Strategy types I've built and validated in production include intraday momentum and trend-following systems, mean reversion and stat-arb frameworks, breakout and scalp architectures with dynamic sizing tiers, options strategies including iron condors, defined-risk vertical spreads, 45-DTE and 0DTE SPX structures with delta-based hedge logic, and GEX (gamma exposure) driven directional frameworks using open-interest-weighted dealer positioning data. It's an area where the signal literature is still thin and where getting the methodology right matters more than most people acknowledge. - Execution Engineering The research-to-live gap is where most systems break. I build execution infrastructure that closes that gap: Interactive Brokers TWS via ibapi with state machine architectures, order ladder geometry, true breakeven calculations with commission and slippage factored in, and hedge math for multi-leg positions. On the NQ/ES side I've built and audited scalp bot execution logic on Tradovate, integrated Databento for professional tick data pipelines with proper validation against reference data, and developed Pine Script strategies and indicators on TradingView for signal prototyping before committing to full implementation. I'm also familiar with NinjaTrader's execution environment and ATM framework. Code I deliver is production-standard: structured logging, explicit edge-case handling, clean separation of signal and execution layers, and documented at a level where the system is maintainable after handoff. - What I Work On: - Strategy Research & Backtesting: QuantConnect/LEAN, Backtrader, Zipline, VectorBT, QuantLib, Python - Execution Systems: IB TWS/ibapi, Tradovate API, NinjaTrader, REST/WebSocket broker integrations, FIX Protocol - Retail Platform Development: MQL4/MQL5 (MetaTrader 4/5), Pine Script (TradingView), NinjaScript (NinjaTrader 8), EasyLanguage (TradeStation), ThinkScript (Thinkorswim), cAlgo/C# (cTrader), AFL (AmiBroker) - Broker APIs: Interactive Brokers, Alpaca, TD Ameritrade, Tradovate, Oanda, Databento - Crypto Platforms & APIs: Binance, Bybit, OKX, Kraken, Coinbase Advanced Trade, Bitfinex, Deribit (options/derivatives), KuCoin, Gate io, Freqtrade, Hummingbot, 3Commas, Pionex, Gunbot, ccxt (unified exchange library) - Statistical Testing: Walk-forward testing, parameter sweep analysis, Monte Carlo simulation, bootstrap inference - Full-Stack Delivery: Data ingestion → signal research → backtesting → live deployment