@ezradesmonds

Composing signal

Back to Work
Quantitative Research / FinTech / AI Research System

FinanceOS

A private financial research and dry-run trading terminal for market analysis, algorithmic signal scanning, and structured trade evaluation.

FinanceOS interface preview

Overview

FinanceOS is a financial research terminal for systematic market analysis using public Binance data. It supports scheduled scanning, signal generation, dry-run trade execution, and performance analytics — all in a structured research environment. This is a research and analysis system, not a live trading platform.

Problem

Traders and quantitative researchers need a systematic, auditable environment for backtesting strategies and evaluating trade ideas without risking real capital.

Target Users

Quantitative researchers, traders, and financial analysts

My Role

Product architect and developer

Product architecture, market-data integration, research workflow design, backend planning, persistence architecture, and frontend terminal experience.

Key Features

  • Public Binance market data integration
  • Scheduled market scanning
  • Algorithmic signal generation
  • Dry-run trade execution and lifecycle tracking
  • Equity curve visualization
  • Performance metrics (profit in percentage and R)

AI System Notes

Provider / Model

Algorithmic signal pipeline with news/sentiment processing; not connected to live brokerage execution.

Data Flow

Binance/public market data, RSS/news providers, and fallback simulators feed the terminal and paper lifecycle records.

Validation

Signals include scoring, regime filters, invalidation notes, and explicit research-only labeling.

Failure Handling

Provider status surfaces disabled/fallback feeds, and live orders stay disabled by design.

Technical Limitations

Research-only system. No wallet, broker, exchange key, or live financial execution is connected.

Pipeline

  1. Public market data and news feeds are ingested into a research workspace.
  2. Rules score trend, momentum, liquidity, BTC regime, risk/reward, and signal freshness.
  3. Signal cards expose entry, stop, TP levels, invalidation criteria, and dry-run lifecycle state.
  4. A dry-run worker tracks paper positions for evaluation without sending live orders.

Challenges & Trade-offs

  • Designing a reliable signal pipeline over public market data
  • Implementing trade lifecycle tracking that accurately represents paper positions
  • Building a research-grade frontend that remains fast with real-time data

Results

Dry-run trading terminal with scheduled market scanning, signal generation, trade lifecycle tracking, and equity curve visualization.

Proof Artifacts

Technologies

Next.js TypeScript Python Binance API PostgreSQL