FIG 0.0 — Origin

About Trader AI

Built by Layer9ine — a custom software studio in Toronto. Trader AI is the internal quantitative trading system powering the Layer9ine personal fund.

FIG 1.0 — The System

What it does.

Trader AI is a fully automated quantitative research and signal generation platform. It monitors crypto, stocks, and ETFs continuously across multiple timeframes, running a suite of strategies through a rigorous regime-gating pipeline — 22,002 candidates tested so far, 112 registered, 107 currently in paper trading.

It was built because off-the-shelf tools don't enforce research discipline. Every strategy that reaches paper trading has passed Monte Carlo robustness validation and an oversight check before deployment — and none of them are trading with real capital yet.

What makes it different: regime-gated selection means strategies only activate in the market conditions they were designed for. Kelly sizing means position risk scales with conviction. And nothing executes without a human in the loop.

FIG 1.1 — Pipeline

SYS 1.0

Data Ingestion

OHLCV from Binance WS. Real-time candle close triggers pipeline.

SYS 2.0

Regime Engine

HMM + volatility regime classification. Outputs: BULL / BEAR / SIDEWAYS / VOLATILE.

SYS 3.0

Strategy Layer

15+ strategies evaluated. Regime-gated ensemble scoring picks winner.

SYS 4.0

Risk & Sizing

Fractional Kelly sizing. Max 5% single position. Portfolio heat check.

SYS 5.0

Execution Gate

Telegram signal sent. Human approval. Jesse 2.0 executes on Binance.

FIG 2.0 — Technology Stack

Built on.

FIG A.1

Strategy Engine

Python · Jesse 2.0 · Binance WS

FIG A.2

Regime Detection

HMM · Volatility Clustering · GARCH

FIG A.3

Risk Management

Fractional Kelly · Portfolio Heat · Max Drawdown Guard

FIG A.4

Infrastructure

Ubuntu 24.04 VPS · PM2 · FastAPI · PostgreSQL

FIG A.5

Human Gate

Telegram Bot · Approval-Required Execution

FIG 3.0 — Contact

Get in touch.

Interested in Layer9ine's custom software services?

layer9ine.com →Email Joel