Alphaware
Writing about what happens when you actually use quant and AI technology.
Anyone who has studied systematic trading eventually discovers a larger-than-expected gap between theory and practice. Methods that work perfectly in textbooks break down when applied to real data, and backtests that seemed airtight produce completely different results in live trading.
Alphaware is where I document what I've learned while trying to close that gap. I write about combining on-chain data, macro indicators, and technical analysis to assess market direction; the pitfalls of applying ML models to financial time series; and the real problems encountered when building AI pipelines like RAG in production.
It leans more toward "here's what went wrong when I tried this" than "here's how to do it." I'm more interested in what actually works in practice than in polished, finished answers.
Tech Stack
Data-driven market analysis combining on-chain data, technical analysis, and macro indicators. Centered on BTC and ETH, I combine on-chain data (exchange flows, whale movements), derivatives indicators (funding rates, OI), and macro signals (DXY, rates) to assess directional bias. Technical analysis is part of the toolkit, but I try not to draw conclusions from a single harmonic pattern.
From AI/ML system building to data pipeline engineering. I write about financial data pipelines, LLM applications, and RAG system construction — the real problems you hit from data ingestion through model serving. Topics include LangGraph, vector DBs, and embedding fine-tuning.
Backtesting methodology, portfolio optimization, and risk management strategies. I cover strategy development and validation methodology: walk-forward analysis, purged K-fold cross-validation, OOS testing, and practical overfitting prevention. The focus is less on building a great backtest and more on validating a strategy that holds up in live conditions.
Newsletter
Get market analysis, quant strategy ideas, and AI & data tool insights delivered to your inbox.
Subscribe →