BrainApp — G&O Studio Case Study

Overview

Confidential client — metrics shared with permission. No strategy details disclosed.

Signals in. Positions out. Nothing in between.

BrainApp is a custom execution engine for a single operator trading USDT perpetuals on Bybit. The system receives signals, applies a multi-layer risk filter, sizes positions, and executes — all in under 42ms from signal receipt to order confirmation.

The operator needed zero operational overhead: no manual order entry, no missed signals during sleep, no emotional override. The system runs fully autonomously with a kill-switch dashboard accessible from any device.

The risk layer was the critical component — hard stop-loss enforcement, daily drawdown limits, position size caps, and automatic trading halt on anomalous market conditions (circuit breaker). 99.98% uptime since launch.

Timeline · 8 weeks

Week 1–2

Architecture & risk design. Signal protocol spec, risk model design, Bybit API capability audit. Decision: Python execution core + Node.js dashboard. Redis for state management.

Week 3–4

Execution engine. Bybit WebSocket integration, order management system, position tracker. Latency optimisation to sub-50ms. Backtested against 18 months of historical data.

Week 5–6

Risk layer. Hard SL enforcement independent of signal, daily drawdown limit, max position size, circuit breaker (halt on -3% daily). All overridable from dashboard.

Week 7

Dashboard. Real-time P&L, open positions, signal log, equity curve, kill-switch. Mobile-responsive. Auth via hardware key.

Week 8

Paper trading & live launch. 2 weeks paper trading, zero incidents. Live deployment on dedicated VPS (Frankfurt, low-latency to Bybit EU). Full monitoring setup.

Technology