Algorithmic Infrastructure

Fluid-dynamic orderbook depth verification.

An automated, institutional-grade trading client observing Binance depth-20 order books in microsecond cycles. We combine macroscopic Optimal Transport (OT) transport maps with Smart Money Concepts (SMC) to isolate verified liquidity pools. Complete paper execution safety.

Launch Terminal terminal Explore Playbooks

M1–M9

Playbook models

—

BTC/USDT price

60s

OT analysis cadence

LiveObserver.sys

Live Pipeline

// Loading SMC playbooks: M1 to M9 compiled successfully.

[00:56:12] ORCHESTRATOR snapshot pipeline connected. Monitoring depth 20.

[00:56:15] OT_ANALYZER Wasserstein W2 cost: 124.92 (spike_rate: 0.12)

Market Structure Narrative BULLISH BIAS

HTF: [BULLISH expansion]. LTF consolidation above HTF Demand OB ($62,850 - $63,050). Waiting for liquidity sweep.

// Monitoring EQL liquidity pools at $62,790...

Click trigger to simulate liquidity sweep

SMC Verification Core

Institutional SMC Models (M1–M9)

The system monitors price action against predefined institutional pattern playbooks. A trade is only registered when all conditions are mathematically satisfied.

Validation Active

Liquidity Pool Raid + MSS Reversal

The flagship model, running live in paper mode. It stalks untaken liquidity pools (equal highs/lows, canonical wick criterion), waits for a raid through the level, and demands a Market Structure Shift before entry. Direction is gated by the 4h dealing-range location; the target is the opposite draw-on-liquidity pool.

Required Validation Steps (Checklist)

✓ Untaken liquidity pool mapped (wick-canon freshness)

✓ Raid: price sweeps through the pool level

✓ Market Structure Shift (MSS) confirms the reversal

✓ 4h dealing-range location gate; target = opposite pool (DOL)

Technological Innovation

SMC narrative meeting fluid transport

We solved the "noise" problem in crypto scalping. By combining macroscopic order book metrics with structure validation, our model screens out invalid sweeps.

water_drop

Optimal Transport Depth Analyzer

Monitors Binance depth-20 orders. We map bid/ask queues as mass densities and compute the Wasserstein-2 distance matrix. High W2 cost represents significant energy spikes, signaling sudden liquidity blocks or smart money spoofing.

▪ W2 Cost metric computed once per minute
▪ Tracks energy spikes and distribution shifts
▪ Signals order book blockages before they show on candles

W2 Mass Transport Cost (Live Profile) W2: 148.9

account_tree

SMC Narrative Engine

Performs top-down structural analysis (1w → 1d → 4h → 1h → 15m → 1m). The system determines the overarching market Bias and Wyckoff Phase, establishing a Watchlist of priority Demand and Supply Points of Interest (POIs).

▪ Weighted bias voting (HTF bias vs LTF bias)
▪ Wyckoff Phase classification: Expansion/Accumulation/...
▪ Automated POI invalidation if price drifts > 5%

Live Context (Matryoshka Event) HTF → LTF Stream

[1w] BULLISH bias confirmed via BOS sweep

[4h] RETRACEMENT phase. Active demand OB located

[15m] RANGING bias. Price approaching target POI

Research Infrastructure

Backtest Lab with live replay

Any listed pair, any window — straight from the dashboard. A four-step wizard checks candle coverage, downloads missing history, runs the exact production M9 engine and streams every setup live over SSE while a TradingView-style playhead replays the run on the chart.

▪ Pair → period → run → report: one wizard, zero CLI
▪ Byte-parity with the research CLI — same engine, same numbers
▪ Replay playhead + setup markers driven by the job cursor

Multipair live: BTC ETH SOL BNB LINK +5

BacktestLab.sys

REPLAY

BTC/USDT · 2025-06-01 → 2026-06-01 · gate: location 0%

▲

▼

Live replay · SSE cursor

+115.5R

Σ 4 pairs · 12 months

151

Setups · M9 location-gate

4/4

Pairs net-positive

Real-time Observability

Fully Automated Paper Execution

Every verified setup lands in the persistent trade journal and is dispatched to Telegram with a rendered chart snapshot of the pattern. The system manages virtual paper positions dynamically, logging entry, stops, and P&L results.

Inconsistency Guard Active Protection

Dynamically penalizes conflicting signals (e.g. going LONG inside Wyckoff DISTRIBUTION phase reduces setup weight by 1.5). Ensures the model trades with structural confluence.

Alert History 24h activity

M9 Pool Raid 00:15:10 · BTC/USDT

LONG RR 3.4

M9 Pool Raid 00:08:44 · SOL/USDT

SHORT RR 2.1

M1 Reversal Yesterday · BTC/USDT

LONG RR 2.3

M6 Turtle Soup Yesterday · LINK/USDT

SHORT RR 1.8

M2 Continuation 2 days ago · BNB/USDT

LONG RR 2.0

Virtual Positions Paper P&L in R

LONG BTC/USDT

Target hit +3.4R

SHORT SOL/USDT

Target hit +2.1R

LONG LINK/USDT

Stopped out -1.0R

LONG BNB/USDT

Watching retest OPEN