Getting Started
NeuroTrader Documentation
Everything you need to understand, operate, and build on the NeuroTrader AI trading platform — from your first paper trade to the REST API.
Sign up, onboard, and run your first AI paper-trading session in about five minutes.
Meet the six engines that analyze the market and how their votes become decisions.
The complete operator manual for running, monitoring, and tuning AI sessions.
Every protection layer, from model agreement to the kill switch.
Programmatic access to your account, trading data, AI sessions, and market data.
Quick answers to the questions everyone asks first.
What is NeuroTrader?
NeuroTrader is an AI-powered cryptocurrency trading platform. In paper mode, its AI engines analyze the market continuously and can generate simulated trades using virtual funds when agreement and risk conditions are met. Controlled live trading, where available, is limited to approved users and remains subject to account, exchange, security, and risk controls. You set the limits; the AI operates inside them; every decision it makes is recorded and explained in plain language.
The platform is non-custodial: in paper mode you use a simulated account with virtual funds. In controlled live mode, where available, assets remain on your own exchange account and NeuroTrader uses user-controlled trade/read-only API permissions. Withdrawal permissions are never required.
How the platform is organized
- System — mission control: platform status, model health, market conditions, and your portfolio performance.
- Trading — the console where you start, monitor, and stop AI sessions.
- Decisions — the audit trail: every trade, skip, and block with its reason.
- Analytics, Signals, Models, Nexus — deeper views into performance, live signals, and per-engine telemetry.
- Settings — your risk controls and trading limits, enforced on every decision.
The Platform Pages guide walks through every page panel by panel.
The core idea: selectivity is the edge
Most of what the AI does is deciding not to trade. A potential trade must clear three hurdles in sequence: enough engines must agree on direction (model agreement), the ensemble must be confident enough (confidence threshold), and all of your risk limits must allow it (risk gates). In quiet or chaotic markets, the correct behaviour is patience — and that is exactly what you will observe.
Weak readings ≠ broken platform
Metrics like "Signal Quality: Weak" describe the market and model environment, which rises and falls with the market. The platform's own health is the separate Platform Status reading. See Reading the Metrics.
Where to go next
- New user? Start with the Quick Start.
- Want to understand the AI? Read The AI Engines and The Decision Pipeline.
- Configuring the platform? The Settings Reference explains every control.
- Building an integration? Head to the API Reference.