Every session. Only A-setups.
Pre-market scan in one click, a graded second opinion in ten seconds, and a hard number on every 'should I?' between the open and the close.
The gap between the last two rows is the entire product.
The 08:45 motion
A watchlist ranked worst-to-best, so the mediocre names stay visible instead of being scrolled past. The session clock above it is real.
Sample A worked example, not a live read. The format is what is on offer here; the numbers belong to a chart from another day.
Three of six are below B and the list says so instead of finding a reason for each. A pre-market scan that returns six opportunities every morning is measuring your hope.
Grade your watchlistSample watchlist so the grades stay comparable. Three of six sit below B, which is roughly what a real morning looks like.
A session with ORIN
Six timestamps. Only one of them involves entering a trade.
- 08:45
Batch-grade the watchlist
Six names, one click, ranked. Two clear B, four do not. The four are not hidden — seeing them scored badly is what stops you arguing yourself into them at 11:00.How the scan works - 09:31
Alert: a watched setup went live
The level you marked pre-market gets hit. The re-grade comes back A− and the plan is already written, so the decision is take-or-pass rather than analyse-then-decide with the candle moving. - 10:40
A second read before adding size
Someone posts a chart in your Discord and it looks like your idea. Run the ticker yourself and get a read built from candles rather than from a stranger's markup, before you add to the position.Scan a symbol - 11:30
The revenge trade grades C — passed, and logged
You are down on the day and this one looks fine. It is a C: mid-range, structure unresolved, and the only thing recommending it is that you want it. You pass, and the pass is recorded so that in a month you can see what your passes were worth. - 16:10
The session journals itself
Every decision, the grade at the time, and one behavioural note. Nothing to write up, which matters because the journal you keep by hand is the one you stop keeping in week three.
The 11:30 entry is the one to read twice. Nothing about it feels like using a product — and it is the only line on this page that reliably makes money.
Intraday calibration, and the overtrading number
A day trader's proof is not the daily-chart hit rate. It is whether the model holds up on the timeframes you actually trade.
Only the last one is a number we can stand behind today, and it is the least interesting of the four. The overtrading figure is the one this page is really about, and inventing it would make the whole argument worthless.
How the calibration worksWhat an intraday workflow uses
AI Chart Scanner
The 08:45 batch and every mid-session re-grade — symbol in, graded read out.
AI chart scannerScreenshot Analysis
Why a read has to be measured off the chart rather than described from it — and where that layer stands.
AI chart screenshot analysisLive AI Studio
When a setup deserves more than ten seconds — chart, score and plan on one screen.
Live AI StudioORIN won't make you faster than the tape
Intraday is where speed claims get made, so here is precisely what this is not.
It is a decision layer, not an execution edge
Analysis takes seconds, not milliseconds. If your strategy depends on being first to a price, nothing here helps and you should not buy it expecting otherwise.
News candles do not care about structure
A headline can erase a level that held all morning. Intraday structure is the most fragile kind, and the tightest stop is the one most easily taken out by noise that means nothing.
A grade is not a fill
The read assumes you get the entry it describes. In fast conditions on thin names, slippage between the graded level and your actual fill can turn a 2R plan into a 1.2R trade.
It cannot stop you clicking
The C-grade at 11:30 only works if you let it. There is no lockout, no cooling-off timer and no broker connection — just the number and whatever discipline you brought.
If most of your volume is gold or the indices, the news-window behaviour matters more than anything on this page — that desk is here.
Questions traders actually ask
The 5m and 15m carry most intraday decisions, with the 1H read as context. The 1m is available but noisier, and the confidence bands widen to say so rather than hiding it behind the same score.
A scan takes seconds, which is fast enough to grade a setup and slow enough that it is not an execution tool. If your edge depends on being milliseconds early, this is the wrong layer of the stack.
That is the main thing it does. Most sessions produce one or two setups worth taking, and the batch scan makes the other four visibly mediocre instead of individually arguable. The value is the trade you did not take at 11:30.
It does not connect to either. You scan by symbol and the plan is yours to carry to wherever you execute. There is no integration to break and no order routing to trust.
Set your account size and cap, and the budget bar shows what each trade consumes and how many losses remain. If you trade a funded account the same maths runs against your firm's published limits instead.