Early access

Algorithmic Trading with AI

From idea to system, no coding.

You decide. AI builds and tests your trading systems.

A Claude Code plugin, the GreaterWaves MQL5 Kit and a step-by-step course to take ideas to validated, portfolio-ready Expert Advisors in MetaTrader 5. The AI does the research, the code and the tests. You decide at every phase.

What early access means

The new course and the plugin are being finished and tested with current students. Join now and you start today with the current program, and get every new part as it’s released — at no extra cost.

Available today

  • The current program: Trading with Generative AI in MT5, Robust Profitable Systems and MQL5 Bot Programming — about 40 hours
  • The Master EA with full, commented source code
  • Private community: questions answered twice a week and a monthly group session, always recorded

You get each part by email when it’s released. Already a student of the program? You receive all of it at no extra cost.

Generating a strategy is easy. Knowing which one to keep is not.

AI can write an Expert Advisor in minutes. Most of them look great in a backtest and fail in live markets: over-fitted parameters, no real validation, costs ignored, no portfolio behind them. Faster code doesn’t fix that. A process does.

Build a system

Seven phases from raw idea to final report, each with a decision. Weak ideas are dropped before any code is written.

Build a portfolio

Compare combinations, split risk by asset family, check correlation on the worst days — for a prop firm challenge or for your own capital with Darwinex.

Take it live

Add the system to a portfolio EA, check it trades exactly like its own backtest, set its risk weight, deploy on a VPS and compare live against backtest.

The AI does the work. You make the decisions.

The AI

  • Researches the idea and the data
  • Writes and compiles the code
  • Runs the tests in MetaTrader 5
  • Writes a report for every phase

You

  • Choose the idea
  • Decide at every phase: continue or drop
  • Approve the risk
  • Decide what goes live
  1. 01

    Edge and design

    Turn an idea into a measurable hypothesis with objective rules.

    Your decision
  2. 02

    Pre-study

    Check that the idea has a real edge, with real costs, before writing any code.

    Your decision
  3. 03

    Build

    The AI writes the system on a multi-strategy MQL5 framework and compiles it.

    Your decision
  4. 04

    Validate

    Rule-by-rule evidence and your own visual test, trade by trade.

    Your decision
  5. 05

    Optimization plan

    One variable at a time, with ranges that make sense.

    Your decision
  6. 06

    Optimize

    Improve only what clearly holds up. A small gain is not a reason to change.

    Your decision
  7. 07

    Final test & verdict

    A final test kept untouched until the end. Portfolio-ready, under observation, or discarded.

    Your decision

How every phase is taught

1. Why

What the phase does and which mistake it prevents.

2. The AI at work

A real recording, edited, with every request and correction visible.

3. The report

How to read each chart and table of the phase report.

4. The decision

Continue or drop — and why. One real system runs through the whole course.

Two destinations, one process

Funding is faster but uncertain in the long run; Darwinex is the long-term path. The course covers both.

Faster, less certain

Funding with prop firms

The quickest way to trade larger capital — but uncertain in the long run, because of how prop firms make money.

  • The firm’s actual rules, program by program
  • Position limit and risk per position
  • A funding score and how many challenges to buy
Long term

Capital with Darwinex

Build an audited track record with your own capital or virtually with Darwinex Zero, and grow it over years.

  • Risk split by asset family
  • The VaR Darwinex would measure, and your estimated DARWIN
  • Return against drawdown, resampled

The rules behind this process run my own audited accounts

The validation and portfolio rules this process enforces are the ones behind these DARWINs. The AI workflow itself is new: I’ll publish the systems built with it as they go live.

5 of 7

ideas I took through this process were discarded before risking a cent. One is ready for a portfolio, one is under observation.

DARWIN GWS

Multi-strategy tactical alpha on liquid global markets

DARWIN GWS
  • Live track record since 2022
  • Max Drawdown -23%
  • Annualized return 8%
  • AuM €100,000

DARWIN UMH

Volatility-adjusted alpha on recurrent market behaviors

DARWIN UMH
  • Live track record since 2025
  • Max Drawdown -9%
  • Annualized return 24%
  • AuM €480,000

DARWIN PWZT

Systematic equity allocation based on cross-sectional signals

DARWIN PWZT
  • Live track record since 2025
  • Max Drawdown -8%
  • Annualized return 21%
  • AuM €360,000

See all live systems →

Past performance does not guarantee future results. Trading involves risk. This content is educational and is not an invitation to invest or investment advice. Data as of September 2026.

Is it for you?

For you if

  • You want to build systematic strategies with a repeatable process
  • You will validate before risking capital
  • You think in portfolios, not single trades
  • You accept drawdowns as part of the process
  • You’d rather drop ten ideas than trade one bad one

Not for you if

  • You want signals or copy trading
  • You expect guaranteed profits
  • You want a bot that “just works”
  • You believe a backtest alone is enough
  • You want to live from trading in a few months

What you need

A Claude plan

For the plugin: a paid Claude plan, billed by Anthropic and not included. Pro is recommended.

MetaTrader 5 on Windows

The process runs its tests in MT5’s Strategy Tester. Mac is not supported for now.

A broker or prop firm account

For demo and live trading. No large capital required.

Time, not code

Progress depends on how many ideas you test. No programming required.

Join early access

Early access

$1,700

One-time payment

  • Everything available today
  • The new course, plugin and kit as they’re released
  • Community and monthly group session
  • All future updates
Join early access

Early access + 1:1 mentorship

$4,000

Flexible payment options

  • Everything in early access
  • 12 weekly 1:1 sessions over 3 months to launch your first portfolio
  • Direct access to me between sessions
  • Max. 3 students at a time
Ask about mentorship

Access is immediate and includes downloadable code, so enrollment is non-refundable once activated.

Frequently asked questions

What do I get today, exactly?

The current program (three courses, about 40 hours), the Master EA with its source code, and the community with the monthly group session. The new course, the plugin and the kit arrive as they’re released, at no extra cost.

When are the new course and the plugin released?

They’re being finished and tested with current students in a closed beta. You get each part by email as soon as it’s ready. I won’t give a date I can’t keep.

Do I need to know how to code?

No. The AI writes and compiles the code on the kit; you check behaviour with a visual test and decide at every phase.

Will I get a profitable system?

No one can promise that. You get a process that drops weak ideas early and documents why the survivors passed.

Which Claude plan do I need?

Pro is recommended. The process runs in Claude Code on your own subscription, billed by Anthropic.

Does it work on Mac?

Not for now: the process runs its tests through MetaTrader 5’s Strategy Tester on Windows.

How is this different from strategy generators like StrategyQuant?

Generators produce thousands of candidates and keep the best-looking backtest. This process starts from one idea with a reason behind it and validates it phase by phase, before and after writing code, keeping a final test untouched until the end.

Can I use it for prop firm challenges?

Yes. The portfolio part scores a portfolio against a firm’s actual rules and tells you the position limit, risk per position and how many challenges to buy.

I’m already a student. Do I pay again?

No. Current students of the program receive the new course, the plugin and the kit at no extra cost.

Refunds?

Access is granted immediately and includes downloadable code, so enrollment is non-refundable once activated. Full terms in the Terms & Conditions.

Build systems that survive.

If you’re ready to test ideas properly and keep only the ones that hold up, the next step is yours.