How ORION Thinks. Step by step.
Probabilistic execution. Constrained decisions. Transparent reasons.
ORION is not a magical AI. It is a probabilistic execution engine with hard guardrails. Every signal you see has survived a deterministic pipeline, a grading system, and a circuit breaker designed to protect capital first.
Not a Bot. A Constrained Decision System.
ORION operates inside three boundaries that never bend. Every candidate setup is judged against all three before it is ever shown to a member.
Probabilistic, Not Predictive
ORION does not predict where price will go. It ranks setups by how often similar institutional footprints have resolved in our favor.
Multi Layer Confirmation
A setup must pass SMC structure, higher timeframe alignment, macro context, and confidence scoring. One layer failing means no signal.
Risk Managed by Default
Every signal carries an SL and TP grounded in structure. If conditions are hostile, ORION refuses to trade. Capital protection comes before activity.
Five Stages. Zero Exceptions.
Every published opportunity walks through the same five stage pipeline. If any stage rejects, the signal is logged as a rejection and never reaches you.
Liquidity Sweep
ORION scans M15 structure for institutional liquidity grabs. No sweep, no setup.
Displacement & BOS
A body close break of structure confirms institutional intent, not a wick.
Macro Defense
Real time macro feeds and event blackouts can veto a technically clean setup.
Confidence Gate
Low confidence ideas are rejected. Duplicate ideas at the same level are skipped.
Grade & Publish
Survivors are graded A+, A, B, or C and routed to the right audience.
Every Signal Wears Its Quality.
Not every valid signal is equal. The grade is calculated from higher timeframe alignment, macro alignment, confidence score, and the number of concerns ORION raised internally.
High Conviction
HTF aligned, macro aligned, high confidence, zero concerns. Rare by design (around 10 to 20 percent).
Strong Setup
The majority of published signals. Clean structure with at most a minor concern.
Developing
Valid but with mixed context. Off by default for members who want only premium grade.
Watch Only
Marginal setups kept internal or shown when developing setups are explicitly enabled.
Guardrails That Refuse to Bend.
Beyond the pipeline, ORION runs a layer of hard rules that can pause the engine entirely. These controls are non negotiable.
Auto Pause Circuit Breaker
After three consecutive stop losses, ORION pauses automatically. No new signals fire until the state is reviewed and cleared. Capital preservation comes first.
Macro Veto
If the fundamental backdrop opposes the technical idea, ORION vetoes the setup. A flawless chart should never fly blind into a macro shock.
Duplicate Guard
If a similar setup is already live at the same level, the new candidate is skipped. No noise, no over allocation.
Low Confidence Rejection
When the confidence score falls below threshold, the signal is rejected and logged. Marginal ideas do not earn a slot.
Every Decision Leaves a Trail.
Every ORION run records what it saw, what it considered, what it rejected, and what it drafted. Candidates, evaluations, rejections, duplicate skips, and final drafts are all captured so behavior can be audited, not guessed.
A stabilization watchlist tracks rejection rate, A+ frequency, signal cadence, auto pauses, and expectancy drift over rolling windows. If the engine starts to behave outside its envelope, the data shows it before a member ever feels it.
Fewer Signals. Higher Trust.
You will see fewer signals than a typical retail bot, and that is the point. Each one carries a visible grade, a transparent reason, and the same guardrails that paused ORION in the past when it was wrong.
For educational purposes only. Not financial advice.