No coding. The AI researches your idea, codes it and puts it through a phased validation process to check it’s robust. You decide what matters: the criteria, the doubts and the verdict.
No hype. No shortcuts. Just process.
AI can write an Expert Advisor in minutes. Most of them look great in a backtest and fail in live markets. Faster code doesn’t fix that. A process does.
Thousands of combinations tested until one looks good. With enough attempts, one always does — by chance.
Parameters tuned until the past looks perfect. The future doesn’t cooperate.
One bot with no portfolio behind it and no real validation.
Every system goes through seven phases. Most ideas are dropped early — and that is the point. The process moves on by itself while the idea meets its criteria, and calls you when there’s something to decide. What used to take weeks of coding and testing now takes hours, or minutes.
Turn an idea into a measurable hypothesis with objective rules.
Pass criterionCheck that the idea has a real edge before writing any code.
Pass criterionThe AI builds the system on a proprietary, portfolio-ready MQL5 framework.
Pass criterionProof that the code does exactly what the idea says, plus your own visual check.
Pass criterionDecide what is worth optimizing, and what is not, before touching anything.
Pass criterionOptimize to understand what truly works and has an edge.
Pass criterionA final test kept untouched until the end. Portfolio-ready, under observation, or discarded.
Pass criterionThen the systems that survive go into a portfolio: for a prop firm, for your own capital, or for capital you manage for others.
The same systems, combined into a portfolio built for where you want to trade.
The quickest way to trade larger capital — but uncertain in the long run, because of how prop firms make money.
Trade your own account, build an audited track record (for example with Darwinex or Darwinex Zero) or manage capital for others, and grow it over years.
These DARWINs, verified by Darwinex and Darwinex Zero, run systems I built before this AI process. Here they are as they stand, each with its drawdown. The systems built with the new process will be added as they go live.
ideas I took through this process were discarded before risking a cent. Two are ready for a portfolio; seven are under observation or held in reserve.
Multi-strategy tactical alpha on liquid global markets
Since June 2026: quantitative short-selling strategy across equities.
Volatility-adjusted alpha on recurrent market behaviors
Systematic equity allocation based on cross-sectional signals
Trend-following strategies in crypto markets with asymmetric risk
S&P500 long-only rule-based strategies that switch to cash to protect capital
Past performance does not guarantee future results. Trading involves risk. Darwinex Zero track records are based on virtual accounts. This content is educational and is not an invitation to invest or investment advice. Data as of September 2026.
9 short lessons, under 40 minutes. A popular strategy put to the test live with AI, a real system taken to a portfolio, and what you can realistically earn.
Free access. No spam. Unsubscribe anytime.
Upload your MT5 Strategy Tester report and replay every real starting point against a prop firm’s rules: probability of passing, expected value per attempt and how long it takes. Your file never leaves your browser.
Open the simulatorFrom idea to system, no coding.
The full process as a Claude Code plugin, the GreaterWaves MQL5 Kit, a step-by-step course and a monthly group session. Join the waitlist and I’ll email you when early access opens.
See the program
I’m José Martínez. I operate systematic portfolios under real capital allocation standards, and my results are publicly audited on Darwinex.
I tried manual trading, automated strategies and strategy-mining software. Most of it failed in real markets. What was missing wasn’t a better indicator — it was a process. This is the one I use.
