Describe your strategy in plain English. We build it, test it against a year of tick data — honestly — and if it survives, you run it live on your own account. No code. No guru. No trust-me.
Fair warning: most ideas don't survive. That's the point.
Right now a human quant runs your autopsy by hand. That's why the verdict takes 48 hours and why we only take a handful a week — and why it's $9.99 while we're still proving this out. The self-serve platform — spec in, verdict in ten minutes — is what the waitlist is for.
Three ideas traders actually trade. The desk runs each one through the full gauntlet right here — and it doesn't pretend they all survive.
Auto-running · click any idea to test it
These five stages are what the desk does to every idea it takes. Today a quant runs them by hand and emails you the report inside 48 hours. The stages themselves don't change when the software takes over — you just stop waiting two days for them.
Talk to the desk the way you'd talk to a quant sitting next to you. It asks the clarifying questions a good engineer would ask.
The AI writes the specification — you sign every rule. Nothing runs that you didn't approve. The spec is the contract; it's hashed and versioned.
Filled at the bid and ask that actually existed — at 14, 27, and 220 milliseconds after your signal. Spread paid. Commission charged. Anything else is fiction.
| signal (UTC) | side | fill | lat | exit | p&l |
|---|---|---|---|---|---|
| 13:42:07.114 | SHORT | @bid 21,847.25 | 27ms | TP | +$298 |
| 14:03:52.881 | SHORT | @bid 21,872.50 | 27ms | SL | −$502 |
| 14:39:10.446 | SHORT | @bid 21,858.75 | 27ms | TP | +$298 |
200 shuffled-world null tests, out-of-sample confirmation, day-concentration audits, parameter-cliff checks. This is the step every other tool skips.
30 days of forward paper proof first — no exceptions. Then the runtime is yours — you run it on your own machine, wired to your own broker via your TradersPost or PickMyTrade webhook. We never touch your money, and we never will. Stage 5 ships with the self-serve desk — today the autopsy ends at the verdict report.
Every day after the session closes, the desk replays your frozen strategy on that day's data and posts the result. If live behavior drifts outside the backtest's cone, you get flagged before your money does. This is what "running your own book" actually looks like.
FEE MATH: $149–267 PER EVAL ATTEMPT · $80–99 RESETS BETWEEN (APEX / TOPSTEP / MFFU PUBLISHED PRICING) · SLIDE TO YOUR HONEST NUMBER
The questions you argue about in Discord at midnight — how many evals to run, when to quit each day, where your stop actually belongs — have exact answers. We compute them from your strategy's own trades, through your firm's own rulebook.
Your verified strategy, simulated through 10,000 funded careers at your firm — trailing drawdown counted tick-by-tick.
| 5 evals in parallel vs 1 at a time | first funded: 2 days, not 12 |
| Expected cost to funded | $412 · ~2.3 attempts |
| Cutting the day off at −$500 | account deaths 40% → 22% |
| "Safe" +$750 daily profit lock | taxes your take −31% |
Eval → funded → first payout, as a computed timeline — before you pay any firm anything.
Your exits, derived from where your trades actually run — not round numbers. Every trade's full journey on tick data, winners and losers.
Plus the warnings: wins-tiny-loses-big shapes · one-magic-number cliffs · stops parked in sweep zones.
Five minutes of conversation. A year of tick data. An honest answer — and if it's yes, a desk that takes you all the way to live.
Order Your Autopsy — $9.99
A quant runs it by hand · verdict in 48h · limited slots each week
Want the ten-minute self-serve version instead? Join the waitlist.
Check your inbox — your payment link and next steps are on the way.