01 · RULE AUTHORITY
The plan becomes executable.
Session windows, max trades, risk limits, re-entry rules and no-go conditions become measurable constraints—not intentions that disappear after the close.
Plan baseline · active
SENTINEL AI · operational trading intelligence
The trader behind the trade.
SENTINEL AI connects execution data, journal context, chart evidence, plan rules and performance history into one operational trading review system—built from your own record.
01 · evidence ingress
Execution record
Entries · exits · size · timing
Sequence locked
02 · evidence ingress
Market evidence
Charts · heatmap · order flow
Context attached
03 · evidence ingress
Journal state
Intent · emotion · reflection
Narrative compared
04 · evidence ingress
Plan authority
Limits · setups · no-go states
Rules enforced
OPERATIONAL REASONING
Not another chatbot. A review authority.
Generic AI can answer a question. SENTINEL AI works inside the performance system, where every answer must survive the trader's own evidence.
Operating premise
Context enters.
A directive leaves.
A directive leaves.
02 · TRADER MEMORY
The review keeps its memory.
Markets, setups, objectives, emotional triggers and recurring behavior remain active across sessions. SENTINEL AI never starts the next review from zero.
Personal context · retained
03 · CONTRADICTION
Words meet evidence.
When the journal reports calm execution but the trade sequence shows urgency, the contradiction becomes a review signal instead of a buried note.
Narrative compared · locked
“We seek partners who are interested in disposing of theater in favor of outcomes.”
Alexander C. Karp · Q1 2025 Letter to Shareholders
CONNECTED INTELLIGENCE
One core. Connected to every part of trading development.
SENTINEL AI does not sit beside the platform. It operates through it—turning every verified input into one coherent development state.
Connected trading development architecture
Select a layer to inspect its authorityPersonal baseline
Trader memory
Markets, style, goals and recurring behavior keep every review specific to the trader.
Feeds the review core
PERSISTENT TRADER MEMORY
The system remembers what the trader forgets.
SENTINEL AI carries the active development state across sessions. Every new review begins with the trader's markets, objectives, triggers and proof target already loaded.
06
Memory fields
120
Sessions sampled
01
Active directive
Active trader dossier
Context state · locked
Memory active
How this changes the review
Review language adapts to the trader's execution rhythm and decision horizon.
DAILY REVIEW ENGINE
Evidence in. One decisive output.
SENTINEL AI reconstructs the session, isolates the controlling behavior and converts the finding into a directive that can be tested the next day.
Wednesday · NQ review
Command record reconstructed
9
Executions
74
Process score
11
Comparable days
01
Evidence
9 NQ executions captured
Three re-entries after the first loss
Best setup appeared inside the plan window
Late-session trade quality declined
02
Interpretation
Self-report and execution diverged
Urgency increased after missed opportunity
Risk remained stable; selection weakened
The same sequence appeared in 11 sessions
03
Directive
Protect the first A-grade setup
No re-entry for 90 seconds after a stop
Restore risk only after confirmation
End the session after two impulse entries
Next-session authority
No re-entry without confirmation.
Generated from 11 comparable sessions and Trading Plan §4.2. Review support only—not trading advice.
CONTRADICTION DETECTION
Patterns become visible before they become expensive.
One trade can mislead. Repeated agreement—or disagreement—between plan, journal and execution establishes the stronger signal.
Journal states
“Calm and selective.”
Execution proves
Four rushed re-entries after missed opportunity.
SENTINEL AI finding
Perception and behavior diverged.
Preserve the pre-market narrative. Remove re-entry authority for 90 seconds after any stopped trade.
Session window
Verified
Max trades
Attention
Daily loss limit
Verified
Re-entry rule
Attention
FULL-CONTEXT REVIEW
Not just P&L. The entire decision state.
SENTINEL AI reads performance beside the evidence that produced it. Screenshots, plan rules, journal entries and execution data stay connected through the final directive.
01 · verified input
Chart evidence
Structure · levels · order flow
02 · verified input
Trading plan
Rules · limits · setup checklist
03 · verified input
Journal
Intent · state · reflection
04 · verified input
Execution
Timing · size · sequence
CLOSED IMPROVEMENT LOOP
Every session strengthens the next.
SENTINEL AI is not a one-time answer engine. It preserves the operational memory required to turn one review into measurable development.
Phase 01
Capture
Executions, market evidence and trader context enter one complete session record.
Complete record established
Phase 02
Compare
Actual behavior is measured against rules, objectives and comparable sessions.
Difference made visible
Phase 03
Isolate
SENTINEL AI identifies the variable with the greatest influence on the result.
Controlling variable exposed
Phase 04
Direct
Analysis contracts into one measurable action with authority for the next session.
One directive assigned
Phase 05
Verify
The following session proves whether the directive changed execution and behavior.
Development verified
Feedback return
The loop stays closed.
Verified evidence updates trader memory.
The next session starts with sharper context, a narrower directive and a proof target already in place.
SENTINEL AI · EARLY ACCESS
Review the trader. Not just the trade.
Connect executions, journal, chart evidence, plan rules and trader memory into one operational review system.