// case study

TradeScope

A full-stack, automated index-options trading platform — built solo, end to end. Real-time data pipeline, risk-managed execution, a live web dashboard, and a cross-platform mobile app, all sharing one backend.

GoReactFlutterPostgreSQLFirebasenginx
// overview

Overview

TradeScope is an automated trading platform for index options. It ingests live market data in real time, evaluates it through a proprietary rule-based signal engine, and executes trades automatically through a broker API — all wrapped in a dedicated risk-management layer.

It runs as an always-on backend with a real-time React web dashboard and a Flutter mobile app for full remote control. The entire system — three codebases, the infrastructure, and the ops tooling — was designed, built, and shipped solo.

// problem

The Problem

Trading a systematic strategy by hand is slow, inconsistent, and emotionally taxing. Opportunities are missed, exits are hesitated on, and risk discipline slips exactly when it matters most.

The goal was a system that could:

  • Act on a rule-based signal the moment conditions are met — without hesitation.
  • Enforce risk limits mechanically, every single time, with no manual override drift.
  • Run unattended and reliably, recovering on its own after any interruption.
  • Be observable and controllable from anywhere, with a safe way to validate changes before risking capital.
// architecture

Architecture

A single Go backend is the heart of the system, with clean separation of concerns: a market-data pipeline, the signal engine, an order/risk manager, persistence, and a REST API. It runs 24/7 and auto-recovers its state on restart.

Two clients consume that one API. The React + Vite dashboard handles live monitoring, interactive candlestick charts, configuration controls, and admin feature flags. The Flutter app mirrors that control on mobile, adding push notifications and analytics. Data persists to a managed PostgreSQL instance, and a real-time notification layer pushes activity and end-of-day reports out over Discord webhooks.

Everything sits behind an nginx reverse proxy on a cloud VM, with auto-renewing SSL on a custom domain, key-based API auth on every endpoint, and a broker OAuth login flow.

// key features

Key Features

Real-time market ingestion

Streams live market data into a proprietary, rule-based signal engine that runs continuously throughout the trading session.

Automated execution

Places and manages orders automatically through a broker API, with a full risk-management layer governing position sizing, stops, trailing exits, and daily limits.

Safe paper-trading mode

A fully isolated simulation mode runs alongside live execution, so behavior can be validated end to end before any real capital is at risk.

Backtesting & analytics

An in-house backtesting engine and performance-analytics module evaluate strategy behavior over historical data.

Full remote control

A cross-platform mobile app gives complete remote control of the system — configuration, monitoring, and live charts from anywhere.

Multi-channel alerts

Real-time notifications for activity and automated end-of-day reports keep the operator in the loop without watching a screen.

// tech stack

Tech Stack

Backend

Go — market-data pipeline, signal engine, order/risk manager, persistence, REST API

Database

PostgreSQL (managed)

Web Dashboard

React · Vite · TailwindCSS · Firebase Hosting & Auth

Mobile App

Flutter · Clean Architecture · BLoC

Infrastructure

Cloud VM · nginx reverse proxy · auto-renewing SSL · custom domain

CI/CD & Ops

Pull-based push-to-deploy · automated backups · uptime health checks

// engineering highlights

Engineering Highlights

  • Built solo across three codebases (Go, React, Flutter) sharing a single backend API.
  • Real-time, fault-tolerant data pipeline with auto-reconnect and heartbeat monitoring.
  • Backend runs 24/7 and auto-recovers its full state on restart — no manual intervention.
  • Seamless database-provider migration with zero data-model changes and no downtime.
  • Safety-first design: every configurable behavior is gated and defaults to a safe state.
  • Pull-based CI/CD with no inbound SSH — build-gated, push-to-deploy, run by a scheduled job on the server.
  • Key-based API authentication on every endpoint, plus a broker OAuth login flow.
  • Cost-optimized cloud deployment running for a few dollars a month.
// outcome

Outcome

TradeScope went from concept to a production system running live, unattended, around the clock — a complete trading operation spanning data, decisioning, execution, risk, and observability, owned end to end by one engineer.

It executes its strategy with mechanical discipline, validates changes safely in paper mode before they ever touch real capital, recovers from interruptions on its own, and stays fully controllable from a phone — all on a cloud footprint that costs just a few dollars a month.

More than the trading itself, the project is a demonstration of full-stack systems engineering: a fault-tolerant real-time backend, two polished clients, and a lean, secure, self-healing deployment — designed, built, and operated solo.