Technical Debt Triage — Prioritized, Evidence Based Backlog
by Leonardo Hermoso
Turn codebase signals into a focused backlog: verified root causes, separate investigation leads, transparent priorities and actionable acceptance criteria.
Around 2007, I was an intern automating options analysis with Excel and VBA, using Mega Bolsa market data. Work done on a calculator started becoming executable rules. Years later, right here on Twitter, I met traders who had theses and needed someone to turn them into code. I wrote, tested and refined Forex strategies. That path led to Elif Capital, where I built the research, validation, risk and execution stack. The code, models and brand were acquired in 2022. I also built DeFi liquidity and hedging systems. At MuscleWiki, I applied mathematical optimization to workout generation. Different domain, same work: turning constraints into decisions. Now I’m building Quantty, starting with CFTC positioning data. One thing keeps becoming clearer, writing code that finds a good-looking result is much easier than figuring out whether that result means anything. In my recent COT research, I couldn’t validate a statistical edge. Finding that out is quant work too. Pushing until the backtest agrees with me would be something else. That’s what I want to share more of here, hypotheses, data, code, execution and the mistakes along the way. Reverse-engineering markets.
by Leonardo Hermoso
Turn codebase signals into a focused backlog: verified root causes, separate investigation leads, transparent priorities and actionable acceptance criteria.
by Leonardo Hermoso
Process a GitHub backlog in dependency-aware waves with isolated worktrees, independent reviews, CI gates and sequential authorized merges.