Your ledger. Reading itself.
Every fill auto-imports from your broker. Setups, regimes, and tags are inferred from the same engine that ranked the trade in the first place. The copilot annotates each entry with stats grounded in your last 90 days — not generic advice.
- <10sfill → card
- 23auto-tags
- 90dstats window
sage ai · Clean execution. Entry 0.4σ above POC, stop respected the value area low — your win-rate on this exact pattern is 73% over 41 prior instances.
Three reasons most journals don’t work.
If your journal is a spreadsheet you forget to open, you don't have a journal — you have an unread document. The Journal closes these as defaults, not features.
- 01manual entry
Journals you stop writing in.
Hand-typed entries collapse on contact with a busy session. By Friday the week is half-blank, and the half that's there is missing the regime, the size, and the stop. The journal you don't keep is not a journal.
0fields you fill - 02generic advice
Coaching that doesn't know your tape.
Most journaling tools tell you to "size down on Mondays" because every account does. We grade your last 90 days against your own ledger, so the annotation is grounded in your prints — not someone else's playbook.
90drolling window - 03lost context
Trades without their tape.
A trade is a tuple of (chart, footprint, regime, narrative). Most journals keep the chart and lose everything else. We capture the whole frame at execution time so the replay is the trade, not a screenshot of it.
5frames per trade
From fill to insight in under ten seconds.
Three stages. None of them require you to type. The journal is what the platform writes when you trade.
- step 00ingest
Broker fill arrives
We connect to Tradovate, NinjaTrader, IBKR, and the major retail brokers via authenticated read-only sessions. Bracket fills land in the journal within ten seconds of the print.
[fill] ES Mar26 · LMT 5,812.25 · 3 · ▴ filled - step 01enrich
Auto-tagged & framed
The same engine that ranked the desk candidate infers setup, regime, session, and risk profile. It captures chart + footprint + tape + value-area at execution time so you can replay the moment, not a memory of it.
tagged · POC bounce · trend·long · session·rth · high conf - step 02annotate
Sage AI grounded note
The copilot annotates each card with stats from your last 90 days — win rate on this exact pattern, your typical hold time, your avg R for this setup, the regimes you bleed in. Citations link to the source trades.
sage · win-rate 73% over 41 prior · avg hold 28m · ↗
Every trade is a document. Five surfaces, one record.
A trade card carries the trade, the frame it was taken in, the plan vs the actual, the inferred tags, and the copilot's grounded read — five surfaces that collapse into one immutable JSON record.
sage ai · Clean execution. Entry 0.4σ above POC, stop respected the value area low — your win-rate on this exact pattern is 73% over 41 prior instances.
real layout · annotated · sample data
- 01Stable trade ID
Every fill gets an immutable JSON document with a ULID. You can link to it, embed it, export it, or audit it.
- 02Chart + frame replay
Captured at execution time: 1m chart, footprint snapshot, value-area, and the active regime band.
- 03Plan-vs-actual
Entry, stop, target, exit, and size — pulled from the broker, not retyped. Diff against your plan is computed automatically.
- 04Auto-tags
Setup family, session, regime, and your custom tags. All inferable, all editable, all queryable.
- 05Sage AI grounded note
The copilot's read on the trade — win rate on this pattern in your ledger, hold-time anomalies, regime fit. Citations to the source trades.
Find where your edge lives. And where it leaks.
Slice the ledger by setup × regime, by session, by tag, by anything. The heatmap below is the default view — a 90-day cohort cut grouped by (setup, regime), colored by average R.
| setup ↓ · regime → | trend·long | trend·short | rotation | vol·rising | vol·crush | shift |
|---|---|---|---|---|---|---|
| POC bounce | +2.4 | +0.6 | -0.3 | +1.8 | +1.4 | -0.8 |
| Demand zone | +2.1 | +0.4 | -0.4 | +1.6 | +1.2 | -0.6 |
| VAL/VAH fade | -0.4 | -0.5 | +1.9 | +0.2 | +1.1 | -0.5 |
| Supply rejection | -0.3 | +2.2 | -0.4 | +1.4 | +0.8 | -0.7 |
| ORB | -0.2 | -0.1 | +0.4 | +1.8 | +0.2 | -1.0 |
| Gap fill | +0.9 | +0.4 | +1.2 | +0.6 | +1.0 | -0.6 |
strongest cell · POC bounce × trend·long · +2.4 R · 41 trades · 73% hit rate
What it isn’t. In writing.
If any of these would be a deal-breaker, better to find out here than after a refund.
- It's not a spreadsheet.
The journal is a queryable ledger of immutable trade documents, not a row-by-row table you maintain by hand. There is no "add row" button.
- It's not generic coaching.
Annotations are grounded in your last 90 days of trades. "Size down on Mondays" is what bad journals say; we say what your ledger says.
- It's not a reporting dashboard.
Reports are downstream. The journal is the source of truth — every chart, P&L line, and tax export pulls from it, never the other way around.
- It's not a copy-trading floor.
Your journal is private by default. You can publish individual trades to your profile or to the leaderboard, but nothing is shared without an explicit toggle.
- It's not coupled to the desk.
If you want to journal trades you took outside Nexural, you can. The desk feeds the journal automatically, but the journal accepts manual entries too.
Questions people actually ask.
If yours isn't here, email sage@nexural.io and we'll add it.
- Tradovate, NinjaTrader, IBKR, Tradier, and TopstepX are first-class on the production tier. Manual CSV upload works for any broker that exports trade history. We add new connectors based on member requests — file one in the contact form.
- Bracket fills land in the journal within ten seconds on the production tier — usually faster. The card is queryable immediately; the Sage AI annotation streams in within thirty seconds.
- Yes. Entries are private by default. You can publish individual cards to your public profile, attach them to a leaderboard entry, or share a read-only link — all opt-in, per trade. We do not train foundation models on private journals.
- Yes. Auto-tags are a starting point. Every card supports custom tags, edits to the inferred ones, and bulk re-tagging from the cohort lab. Your tag schema is yours.
- It’s narrower and more honest. A real coach can read your psychology and your context; the copilot can only read your ledger. Use it for the “what does my data say?” layer, not the “what should I do tomorrow?” layer.
- Yes. CSV export of all closed positions with P&L, fees, and broker mapping. Pro and Automation accounts also get a mark-to-market export and a Form 8949 / Schedule D-friendly format for US filers. Consult a tax adviser before filing — we’re not one.
- Yes. Bulk CSV import will backfill cards. Auto-tagging runs on backfill at the same fidelity as live, but the chart-and-frame replay is only available for trades captured live (we don’t retroactively reconstruct the tape).
Stop typing. Start reading what you trade.
The Trade Journal is included on every paid plan. Connect a broker once, and the next fill writes its own card. If it’s not making your reviews sharper, full refund inside the 7-day money-back window.