ProfitX — Algorithmic Trading Engine

Technologies

TypeScript 90% JavaScript 5% Dockerfile 5%

Topics

trading-bot algorithms automation risk-management postgresql prisma typescript

An automated algorithmic trading engine and order execution pipeline engineered with a finite state machine (FSM) architecture entirely in TypeScript. The system incorporates real-time technical analysis (dynamic RSI and peak/trough detection), market volatility filters, and strict capital preservation guardrails (automated stop-loss, daily drawdown limits, and slippage bounds). Designed with deterministic transaction lifecycles to prevent duplicate orders during high-volatility spikes, coupled with transactional audit logging in PostgreSQL via Prisma ORM and a live telemetry dashboard.

Technical Highlights

Mathematical Indicators

Pure native TypeScript implementations of Bill Williams Fractals (5-candle sliding window) and the Chaos Oscillator (wick pressure), eliminating third-party packages.

Risk State Machine

Automated money management engine utilizing Martingale and Soros state machines with strict risk boundaries (Stop Loss, Stop Win, max exposure).

CQRS Signal Auditing

Robust audit trail mapping incoming signals to order execution logs using correlation and causation IDs for complete CQRS traceability.

Reverse Mirroring

Low-latency synchronization algorithm executing trades on Avalon (via SDK) and mirror-replicating them conversely on ProfitX with reverse direction.

Architecture Case Study

Production codebase is confidential. Public architecture study available above.