If you don't know me, I'm a maintenance tech in Kansas, and for the past 2(ish) years, I've been building trading tools, AI ProtoTypes, and a few dev tools.
The most useful thing I've shipped so far is small: a compounding simulator.
🔗
sfti-ai.org/calc/calc.html
Most calculators add the expected value of each trade to your balance. That's the arithmetic mean, but compounding pays the geometric one. Below a 100% win rate the two diverge, and the gap grows with position size.
Mine plays out 400 simulated paths and shows the median, the 10th–90th percentile band and the naive projection side by side. Fees are charged on both sides, and it reports the share of paths that end below your starting balance.
Behind it is a larger autonomous signal engine running on hardware in my house. It's alpha with no track record yet. I cleared its history on purpose, and its dashboard renders empty rather than show a win rate I can't back.
If you work in quant, risk or simulation, I'd like to know where the methodology is weakest.
#FinTech #AI #Stocks #BuildInPublic
P.S. Feel free to check out the other systems and components available.