The pipeline, in public

How Whalur's AI finds trades.

Six steps, left to right. An AI writes an idea about how the market behaves, we test it against a decade of daily bars, only the ones whose worst-case still makes money survive, and the survivors are watched live and traded on paper so you can see whether they keep working. Every number below is read from the live research database right now — no login, no sales deck.

Step 1

AI proposed

An AI writes new ideas about how the market behaves, in plain English, and hands each one to the test engine.

—

ideas proposed so far

New in the last 30 days —

Step 2

Tested

Each idea is replayed against years of real market history, trade by trade, so it is judged on what actually happened and not on a hunch.

—

simulated trades measured

Test runs in the ledger —
Stocks covered —
Years of daily data —

Step 3

Validated

Most ideas fail, and we say so: only the ones that still make money at the bottom of their confidence range are kept as winners.

—

validated winners

Still being tested —
Waiting to run —

Step 4

Live signals

Every few minutes we scan the whole universe for those exact conditions, and when one fires it becomes a dated, published signal.

—

active signals right now

Last scan —
Mechanisms watched —
Detections today —

Step 5

Paper portfolios

Paper trading — no real money

Signals are traded in paper portfolios with fake cash, so the return you see is a live scorecard rather than a claim about anyone's account.

—

combined paper return

SPY over the same window —
Active paper portfolios —
Open paper positions —

Step 6

Track record

Every finished trade is scored against what the research predicted, and the tally stays public whether it flatters us or not.

—

completed trades scored

Average return per trade —
Win rate —

Loading live numbers…

See the winners.

The counts are public. The entry, stop and target that make a mechanism tradable come with a free account.

Pricing

Free during the research preview. No card required.